Angular displacement patterns of leading and trailing limb joints during galloping in monkeys.код для вставкиСкачать
American Journal of Primatology 22:227-239 (1990) RESEARCH ARTICLES Angular Displacement Patterns of Leading and Trailing Limb Joints During Galloping in Monkeys JOEL A. VILENSKY, MARSHA MOORE-KUHNS, AND ANN M. MOORE Department of Anatomy, Indiana University School of Medicine, Fort Wayne The movement patterns of the joints of the right hind limbs and forelimbs of four vervet and one squirrel monkey were compared during right- and left-lead gallops. Although the overall displacement patterns were similar under both conditions, comparisons of the inflection points in the curves yielded consistent differences in the joints at various phases in the cycle. Some of these differences appear to reflect differing mechanical conditions, whereas others seem to represent variation in muscle activation patterns. Although some of the differences in the joints under the two conditions were consistent with those reported previously for other quadrupeds, others were not. This may reflect that different quadrupeds use slightly different biomechanical strategies during galloping. Key words: primates, locomotion, gait, hind limb, forelimb INTRODUCTION Galloping is an asymmetrical gait in which each homologous pair of limbs (fore- or hind) acts as a unit to propel the animal forward. When the forelimbs (or hind limbs) strike the ground, one of the limbs typically makes contact shortly before the other. The first limb to strike the ground is termed the trailing limb and the second the leading limb [Hildebrand, 19771. Although this terminology appears to be inconsistent, it is based on spatial rather than temporal relationships. That is, while both limbs are on the ground, the second to contact the ground is positioned in front of (i.e., leading) the first to contact the ground. Because of this apparent inconsistency, some prior reports on galloping have reversed the terms [e.g. Miller & Van der Meche, 1975; Miller et al., 19751or have used different terms such as first and second limbs [Cohen, 19791. For this report, we chose to retain the classical terminology. It is intuitively obvious that, because galloping is an asymmetrical gait and because the homologous limbs do not strike the ground simultaneously, the conditions under which each limb contacts the surface differ. That is, the body and the other limbs are in a somewhat different orientation when, for example, the right forelimb contacts the ground compared with when the left does. For this reason as well as because the nervous system may use the leading and trailing limbs slightly Received for publication January 29, 1990; revision accepted J u n e 14, 1990. Address reprint requests to Dr. Joel A. Vilensky, Fort Wayne Center for Medical Education, 2101 Coliseum Blvd., East, Fort Wayne, IN 46805. 0 1990 Wiley-Liss,Inc. 228 / Vilensky et al. differently to achieve efficient propulsion, it is reasonable to expect variation in the joint displacement patterns of leading and trailing limbs. The purpose of this study was to determine if such variation exists in monkeys and, if so, to describe it, analyze it functionally, and compare the results with the limited data available for other mammalian taxa. Regarding other taxa, for cats, Miller and Van der Meche  presented graphs of the angular displacement patterns of the forelimb joints of a single cat when the limb was in trailing and leading positions, but the authors did not discuss any differences in joint angular values apparent in the data. The actual graphs are small and very difficult to use for comparative purposes because all the joints are plotted on the same set of axes. In a second paper, Miller et al.  presented somewhat better graphs for the elbow and scapula joints as well as for the hip and knee joints during trailing and leading conditions. The authors noted, based on these graphs, again of a single cat, that the hind limb and forelimb of the leading and trailing limbs have distinctive features. No quantitative analysis of angular parameters was presented. For dogs, Tokuriki  stated that the joint displacement patterns are only slightly different for the leading and trailing limbs. This evaluation was based on a single animal. The most comprehensive analysis available of differences in limb movement patterns during galloping is for rats. Cohen [ 19791examined in detail the forelimb joint displacement patterns of rats using X-ray cinematography of five animals running on an exercise wheel. Unfortunately, although graphs are presented that display notable differences in joint movement patterns, there is no tabular material detailing how consistent these differences were among the animals. In addition, the graphic comparison of right- and left-lead galloping use data from two different animals, and there is no assurance that both animals were running at the same speed. Thus, although Cohen’s [19791 study clearly indicates pronounced differences in certain joint patterns, there is no indication of how consistently these differences would appear in individual animals. Furthermore, no data are presented for hind limb joints. For monkeys, the only available data regarding differences in trailing and leading limb joint displacement patterns is an estimate by Larson and Stern  that, during swing, the trailing forelimb in vervet monkeys undergoes approximately half the excursion of the leading limb. To collect data for the present study, we examined the lead author’s extensive collection of films of galloping vervet and squirrel monkeys [see Vilensky & Patrick, 1985; Vilensky et al., 1988, 19903 to determine if any of it was suitable for a highly controlled comparison of leading and trailing limb joint displacement patterns. Specifically, we looked for animals that switched leads during galloping at a set speed, i.e., within a filming trial. We also required that each analyzed stride not be part of a transition. Although these conditions were not met by many animals, we did find sufficient data to perform a detailed study. It should be emphasized that, since we always filmed the animals’ right sides, all reported data are for right limbs. Thus this study controlled for any potential asymmetry in limb movements of the two sides of the body as well as for speed. MATERIALS AND METHODS All the film used for this study was taken under similar conditions with the camera operating at a rate of 100 frameslsec except for one squirrel monkey trial, in which the camera operated at 200 frameslsec. Prior to filming, the animals had been trained to locomote at a variety of speeds on a treadmill that was enclosed Galloping in Monkeys i 229 Fig. 1. Sketch of a vervet monkey illustrating the joint angles measured for this study. A, wrist joint; B, elbow joint; C, shoulder joint; D, arm angle (anterior angle between arm and horizontal); E, hip joint; F, thigh angle (anterior angle between thigh and horizontal); G, knee joint; H, ankle joint. within a plexiglass cage. The lens of the camera was positioned 6 m from and perpendicular to the long axis of the treadmill, and the animals ran so that their right sides were closer to the camera. As noted above, the films were examined to determine if any of the animals changed either their forelimb or their hind limb leads for a single speed on the same day (i.e., within a trial). For analysis, we required that there be a minimum of two strides with the noted lead sequence within a trial, but the two strides did not have to be consecutive. When possible, we used trials that had a t least three strides with the desired lead sequences. It is also important to emphasize that, when we chose strides that had forelimb lead changes, we did not control for hind limb leads. We assumed that, within a gallop (when each homologous pair of limbs acts as a unit), the displacement pattern of the forelimb joints (in the sagittal plane) would not notably be affected by the lead pattern of the hind limbs. The same applies when we chose strides that exhibited hind limb lead changes (i.e., for those strides, we did not control for forelimb leads). When more than three strides with a particular lead sequence were available within a trial, we chose three to analyze based on the following two criteria: limbs clearly visible and the monkey showing good posture on the treadmill (no crouching or other unusual behavior). For those strides we chose to analyze, an experienced technician used a cursor to input the estimated positions of the following joint center locations into a computer: metatarsophalangeal, ankle, knee, hip, shoulder, elbow, wrist, and metacarpophalangeal. From these x-y positions, the computer determined the ankle, knee, hip, shoulder, elbow, and wrist angles as shown in Figure 1. As indicated, the shoulder and hip angles were determined from 230 I Vilensky et al. the joint center locations defining those angles (e.g., for the hip angle, the knee, hip and shoulder joint centers; the shoulder angle was defined as the anterior angle so that the same terminology could be applied t o movements at the hip and shoulder). In addition, we also defined thigh and arm “angles” that were simply the relationship of that limb segment to the horizontal. It should be emphasized again that, in all cases, only the movements of the right limb were analyzed. Thus this study involved determining how this limb changed its movement pattern when it was either trailing (i.e., left limb was leading) or leading. For the forelimb and hind limb strides, a stride was considered to begin with right touchdown and end at the frame preceding the next right touchdown. For the forelimb lead changes, data from four animals were available. Data for animal 1 (Al), a 2.3-year-old female vervet monkey, were available for two speeds, 2.58 and 2.81 m/sec. A2, an adult male vervet, exhibited forelimb lead changes at 2.58 m/sec. A3, a 2- year-old female vervet, showed a forelimb lead switch a t 2.81 mlsec. A4, an adult female squirrel monkey, showed forelimb lead changes at 2.28 m/sec. In addition, we used data for this animal for which there was about an 8% variation in speed for the forelimb lead changes, specifically, 1.8 and 1.95 misec. This occurred because the treadmill was still accelerating when one of the lead changes occurred. We used these data despite the variation in speed so that we would have a second set of data points for this animal, our only representative of this species. For the hind limb, data were available for A2 at 2.58 misec and for A4 at 2.28 and 2.58 m/sec. In addition, data were available for one other vervet monkey (A5), an adolescent female (exact age unknown), for 2.28 and 2.81 m/sec. The following number of strides were available for each animal: A l , three strides for both types of forelimb leads at both speeds; A2, three strides for both types of forelimb leads, three strides for left hind limb lead (LHL), and two strides for right hind limb lead (RHL); A3, three strides for left forelimb lead (LFL) and two for right forelimb lead (RFL);A4, three LFL strides at 1.95 d s e c and two RFL strides at 1.8m/sec, at 2.28 m/sec three of both types of forelimb and hind limb lead strides, and at 2.58 m/sec three LHL strides and two RHL strides; and A5, three of both type of hind lead strides a t both speeds. Once each stride was digitized and the angles computed, plots of the angles were printed. From these plots, the values of the joints at the major flexionextension (or extension-flexion) turning (inflection) points were determined. Usually, these points were the highest or lowest before a definitive trend in the reverse direction was apparent (see Fig. 2). We applied the traditional Phillipson nomenclature [see Wetzel and Stuart, 19761 to these inflection points. Briefly, in this system, for a typical joint (e.g., ankle in Fig. 2), an epoch of flexion (yield) that is designated the E2 phase begins at touchdown. Sometime later during stance, the joint begins to extend, beginning the E 3 phase. Thus we called the inflection point between the two, the E2-E3 transition. During late stance, a typical joint then begins to flex again, beginning the F phase, with an E3-F transition between the two. During the F phase, the foot loses contact with the ground and the swing period begins. Finally, later during swing, the joint begins to extend again, beginning the E l phase. The El-E2 transition then occurs a t touchdown of the foot. It is noteworthy that not all the joints analyzed followed the Phillipson pattern. Specifically, the arm, shoulder, thigh, and hip joints did not show an El-E2 or an E2-E3 transition because they did not yield (flex) a t touchdown (no E2 phase). For these joints, we defined an E2-E3 transition as occurring at touchdown even though no actual inflection point occurred at this time (see hip in Fig. 2). This was done so that the touchdown angles could be compared across the animals. The wrist did not show an E l phase and therefore did not exhibit F-El or El-E2 P 2c 80 120 0 195 50 50 PERiEh7 OF CYCLE E2-E3 75 75 103 103 BO 00 25 50 75 103 25 50 PERCENT OF C Y C S 75 100 30 25 50 75 P i a t f L n OF ~ C L E HI? Fig. 2. Displacement patterns for the noted joints of the right limbs of A2 for locomotor cycles (beginning with touchdown) during which the right limb was leading (solid lines) and the left limb was leading (right limb trailing; dotted lines). In addition, the inflection points on the curves are labeled according to the nomenclature describled in the text. For the forelimb graphs, lift-off occurred a t 47% of the cycle, whereas, for the hind limb graphs, it occurred a t 44% of the cycle. RFL = right front lead; LFL = left front lead; RHL = right hind lead; LHL = left hind lead. 25 25 W?’S 100 232 I Vilensky et al. transitions. However, we did define the wrist touchdown angle to be equivalent to the El-E2 transition. For all digitized strides, we determined the values of the joints a t the described inflection points. We then computed mean values and standard deviations for each animal for each set of strides associated with a particular lead sequence. Finally, because this study was based on a small number of animals, presentation of the data is confined to descriptive comparisons using means rather than conclusions based on inferential statistical procedures. RESULTS Forelimb Joints Figure 2 depicts the joint displacement patterns for A2’s right hind limbs and forelimbs, during both right and left lead conditions, beginning at limb touchdown. For the forelimb graphs, lift-off of the limb occurred at approximately 47% of the cycle, whereas, for the hind limb graphs, lift-off occurred at about 44% of the cycle. Clearly, there were no gross differences in displacement patterns for the right limb (fore- and hind) whether it was leading or trailing. There were, however, notable differences in some of the inflection point values for some of the joints. These differences, as well as those for the other animals, are graphically depicted in Figures 3 and 4. Figure 3 depicts the mean forelimb joint values at the noted inflection points for all the angles for all the animals. Each type of symbol in the graph corresponds to a different animal, with the filled symbols being associated with the LFL condition (i.e., right limb trailing) and the open symbols with the RFL condition. Standard deviations are also indicated (upward bars for the LFL condition and downward bars for the RFL condition). Two sets of similar symbols refer to the two speeds used for that animal, with the slower speed to the left. For the El-E2 transition (Fig. 3A), neither the wrist nor the elbow exhibited any consistent changes across the animals. At the E 2 4 3 transition (Fig. 3B), the wrist was generally more flexed in the LFL condition, whereas the elbow joint did not exhibit any consistent changes across the animals. For the arm, across the three vervet monkeys (Al-A3; circles, triangle, and square), when the limb was trailing, it tended to be less protracted a t E2-E3 (touchdown) than when it was leading. The reverse, however, tended to be true for the squirrel monkey (A4; diamonds). Differences in angles were also clearly apparent for the vervet monkey shoulder angles, whereas the squirrel monkey had similar shoulder angle values across both conditions. For the elbow at E3-F (Fig. 3C), all the vervets showed slightly higher values (greater extension) during right leads, but this was not true for the squirrel monkey. The arm angle at E3-F showed a marked consistent difference across all the animals. The arm was always more retracted (extended) at E3-F when the left limb was leading. For the shoulder, the reverse was true, with the angle being consistently greater (more retraction) when the right limb was leading. Only the shoulder showed a consistent difference at the F-El transition (Fig. 3D). The angle was always less (more protraction) when the right limb was leading. Hind Limb Joints For the hind limb joints, there were no consistent differences at the El-E2 transition (Fig. 4A). At the E2-E3 transition point (Fig. 4B) the thigh showed consistent differences, being notably more protracted when the right hind limb was leading. A similar but less marked trend was evident for the vervet monkey hip angles, but this was not true for the squirrel monkey. Galloping in Monkeys / 233 A EI-EZ WRIST FCBOW 220 r 170 120 L 120 T L ARM B. E2-E3 SHOULDER 100 r v? I Y g n 130 120 110 100 90 80 70 - - 6 @2 " 8 0- - Q 8 a i 70 - - 9 0 - T 130 4 90 - 50 40 30 - 30 1 170, 110 I Fig. 3. Mean angular values for each of the noted forelimb joints at the El-E2 (A), E2-E3 (B), E3-F (C), and F-El (D) inflection points. The circles correspond to Al, the triangles to A2, the squares to A3, and the diamonds to A4 (squirrel monkey). Filled symbols correspond to the LFL condition and open symbols to the RFL condition. Where two pairs of similar symbols are presented, each pair corresponds to a specific speed for that animal, with the slower speed on the left (see text). The bars indicate one standard deviation (directed upward for the LFL condition and downward for the RFL condition). For the E3-F transition (Fig. 4C), there were consistent differences for all the joints. For the ankle, the value was always higher (greater plantarflexion) when the right limb was leading. The knee was also always more extended when the right limb was leading. The thigh and hip were both notably more extended (retracted) when the left limb was leading. At the F-El transition (Fig. 4D), the ankle showed no consistent differences among the monkeys. For the vervets, the knee was more flexed when the right limb was leading. For the thigh, the segment was always less protracted when the limb was trailing. This was also true for the vervet monkey hip joint but not for the squirrel monkey. Effects of Speed Speed did not consistently affect the relationships between the joints when the limb was in a leading or trailing position. In general, similar differences were apparent at both speeds for the animals for which data were available for more than one speed. 234 I Vilensky et al. A El-EZ 150 KNEE ANKLE r HIP THIGH 50 100 * i 0 0 “ 70 A 60 180 10 w W CT 150 160 V - v 140 - V * * A A 130 120 - Fig. 4. 50 - i70 160 - y 90 I * 0 ’ ~ 140 I30 1 20 1 160 - 140 - 140 - 120 20 120 110 100 - - v v 100 - ~ 100 ~ A ~ 80 ~ Q 8060 - v v i s + a q o Mean angular values for each of the noted hind limb joints at the El-E2 (A), E2-E3 (B),E3-F (C), and F-El (D) inflection points. The downward-pointing triangles correspond to A5, the upward-pointing triangles to A2, and the diamonds to A4. Filled symbols correspond to the LHL condition and open symbols to the RHL condition. Where two pairs of similar symbols a r e presented, each pair corresponds to a specific speed for that animal, with the slower speed on the left (see text). The bars indicate one standard deviation (directed upward for the LHL conditon and downward for the RHL condition). DISCUSSION To facilitate the discussion of the data presented in the previous section, Figure 5 depicts tracings of A2 during all four conditions. These tracings were taken from alternate frames of the original films (thus each tracing is separated from the next by .02 sec). Each tracing is numbered consecutively and the tracing that is associated with each of the transition points is noted for each joint. Forelimb Joints For the vervet monkeys, the greater arm and shoulder protraction with a right lead a t E2-E3 is associated with the fact that, under this condition, the fore part of the body is already supported by the trailing limb, so the leading limb can outstretch further to increase stride length. This is clearly seen in tracings No. 1 in Figure 5A. The slightly greater wrist dorsiflexion with a left lead a t the E2-E3 transition point may be associated with the fact that, because the right was the first limb t o strike the ground, it showed an increased “yielding” or shock-absorbing (E2) phase. The tendency toward greater elbow extension at E3-F with a right lead may be associated with the limb (being the last to leave the ground) exerting more of a thrust as it does so. The greater retraction of the arm a t E3-F Galloping in Monkeys I 235 when it is trailing probably relates t o the other limb still supporting body weight and thus allowing a greater degree of excursion. The difference in shoulder angles at E3-F, as is apparent from the tracings, relates to differences in the elevation of the hind part of the body as well as to differences in arm protraction (see above). The differences in the shoulder values a t the F-El transition occur for the same reason as described above for the E2-E3 transition (see Fig. 2). Hind Limb Joints The thigh at touchdown (E2-E3) is more outstretched (more protracted) when the limb is leading. This is obvious in comparing tracings No. 1 in Figure 5B. Apparently, the increased protraction of the leading limb is permitted because the limb is striking the ground while the contralateral limb is supporting much of body weight. This explanation also applies to the vervet hip at E2-E3. The difference in ankle plantarflexion a t the E3-F transition seems clearly to be related to the fact that, when the right limb is leading, it is the last to leave the surface and therefore, at lift-off, it imparts a final substantial thrust (compare tracings No. 9). The increased extension for the knee when the right limb is leading would similarly be associated with this final thrusting action. It is clear from the tracings that this thrusting raises the hind section of the body as well as propels the animal forward. The increased thigh and hip extension at E3-F when the limb is trailing is probably associated with increasing stride length (see above). The increased flexion of the leading knee at the F-El transition appears to be associated with the increased elevation of the hind portion of the body. This probably slightly increases the swing duration of that limb. The difference in protraction of the thigh (hip) a t the F-El transition occurs for the same reason as at the E2-E3 transition (touchdown); the leading limb can reach farther forward to increase the stride. Comparisons Between the Hind Limbs and Forelimbs At touchdown, the leading limb (fore- or hind) is more protracted, which, as noted above, is probably related t o increasing stride length. At E3-F, the ankle was more extended (plantarflexed) when the limb was leading. We suggested that this increased plantarflexion was related to a final push-off by the leading hind limb. In contrast, the wrist did not show consistent E3-F differences. This finding suggests that the wrist does not provide much propulsive force. The elbow and knee both tended t o be more extended at E3-F when the limb was leading, which presumably is associated with more power at lift-off. On the contrary, the arm and thigh were both more extended (retracted) at E3-F when the limb was trailing. This probably acts to increase stride length (see above). At F-El, the knee showed a prominent difference, whereas the elbow did not. We suggest that this relates t o the elevated position of the hind portion of the body at this stage (see above). The fact that the arm did not show a notable difference at F-El, whereas the thigh did probably reflects that this turning point was often ill-defined for the proximal joints (see Fig. 2). However, the hip and shoulder tended to be in accord for the F-El point, with the limb being more protracted when it was leading. ComparisonsWith Prior Reports As noted previously, for vervets, Larson and Stern  estimated that during swing the trailing forelimb undergoes approximately half the excursion of the leading forelimb because the trailing limb ends swing phase in a less protracted state than the leading limb. We did not measure “limb excursion” during swing, but our data do indicate that the trailing arm a t touchdown is less protracted than Ankle - El E2 (1) E2 - E3 (5) € 3 - F (9) F - E l (15) Arm Thigh E2 - E 3 (1) E 3 - F (10) F - E l (21) Arm E2 - E 3 (1) E 3 - F (11) F - E l (20) E2 - E3 (1) E 3 - F (13) F - E l (21) Hip E2 - €3 (1) E 3 - F (9) F - E l (19) - Shoulder E2 E3 (1) E 3 - F (11) F - E l (20) - Shoulder E2 € 3 (1) E 3 - F (12) F - E l (21) Ankle El - E2 (1) E2 - €3 (5) € 3 - F (gj F - El(15) E2 E3 (5) € 3 - F (10) F - E l (14) - Knee E l - €2 ( 1 ) Fig. 5 E 3 - F (10) F - €1 (21) Thiah E2 - €3 (1) E2 - E 3 (1) E 3 - F (8) F - E l (18) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - - - - El - E2 (1) E2 - E 3 (5) E 3 - F (9) F - E l (18) El - E2 (1) E2 - E 3 (4) € 3 - F (8) F - El (14) Elbow Wrist El - € 2 (1) E2 - E3 (6) E 3 - F (12) - E2 (1) - E 3 (4) F (10) - El (16) El E2 E3F Wrist El - E Z ( 1 ) E2 - E3(6) E 3 - F (11) Galloping in Monkeys I 237 the leading limb (Figs. 3B, 5A). This is at least in accord with Larson's and Stern's observation. Data from Miller et al. [19751 for the cat elbow indicate major differences for the E2-E3 and F-El transitions. In both cases, there was more flexion in the trailing limb (it is important to note that we have reversed the terminology Miller et al. present in their Fig. 4 because they used leading and trailing in a temporal rather than spatial sense). Our results do not indicate any consistently greater flexion for the elbow at the E2-E3 or F-El transitions for the trailing limb. Regarding the hind limb, the hip joint of the cat showed no notable differences when the limb was in the trailing or leading position [Miller et al., 19751. For the knee, there was markedly greater flexion at E2-E3 and more extension a t E3-F in the trailing limb. The lack of differences for the cat hip joint is surprising compared to the monkey data, although it is possible that variation in measurement techniques accounts for the differing results. For the knee at E2-E3, the monkeys did not show consistent differences, but at E3-F, when the limb was trailing, contrary to the case in the cat, the knee showed less extension than when it was leading. For rats, Cohen [19791 found that the elbow joint differed in trailing and leading limbs at E3-F, F-El, and particularly El-E2, whereas, at E2-E3, the elbow had similar values. At E3-F, the rat elbow was more extended in the leading limb; at F-El, the elbow was more flexed in the trailing limb; and, at El-E2, the trailing elbow was much more extended. The similarity at E2-E3 and the differences at El-E2 resulted in the range of yield (E2 phase) being much greater in the trailing elbow. In comparison, for the monkeys, we also found greater extension in the leading limb a t E3-F but no consistent differences at F-El or El-E2. Interestingly, although the vervets did not show greater yield range (E2) at the elbow when the limb was trailing, the squirrel monkey did (about 12" greater). Cohen El9791 also measured the humerus angle of the rats, which is the same as our arm angle. For this angle a t the E3-F transition point, the trailing limb had notably more retraction than the leading limb. In the monkey, the trailing limb also showed more retraction. Neural Control and Functional Anatomy The data presented here demonstrate conclusively that leading and trailing limbs in monkeys do not undergo identical movement patterns. Although the differences may reflect primarily the differing mechanical conditions associated with the overall positioning of the body, in some cases, e.g., increased ankle plantarflexion in the leading limb a t E3-F, there may be controlled changes in muscle activation patterns in order to maximize efficiency of force production. Accordingly, Cohen and Gans [19751 showed differences in rat triceps muscle activation sequences for leading and trailing limbs. More directly related to the present study, Larson and Stern El9891 found differences in the pectoralis major and latissimus dorsi muscles for the leading and trailing forelimbs of vervet muscles. Specifically, they reported an increase in the activation of the caudal part of the pectoralis major and latissimus dorsi during late swing phase in the trailing limb. The authors associated this increase with the reduced protraction of the trailing Fig. 5. Tracings of A2 taken from alternate film frames (.02 sec apart) during RFL and LFL strides (A) and during RHL and LHL strides (B). Each set of tracings represents a complete locomotor cycle. For each set, the location of the various inflection points for each joint are indicated by referring to the numbers located beneath each of the tracings. 238 I Vilensky et al. limb a t touchdown (see above). Thus these muscles are described as “slowing down the swinging limb.” The EMG data and our data indicate that there is consistent asymmetry in the neural control systems of the two sides during galloping. Perhaps it is sensitivity to the amount of asymmetry in the system that is one cause of frequent lead changes in quadrupeds. Other suggested causes are “whim” [Howell, 19441, fatigue [Dagg, 19771,and prevention of excessive lateral excursions of the center of gravity [Wetzel et al., 19771. The facts that the vervets and the squirrel monkey were sometimes not consistent in the differences found for the leading and trailing limb joints and that there were differences among the monkeys, cats, and rats suggest that variation in body design may result in slightly different galloping strategies. Accordingly, Jenkins  stated that there is not a single mode of posture and locomotion among terrestrial mammals. In addition, the fact that primates may be more dependent than typical quadrupeds on their hind limbs for propulsion may also account for some of the differences [see Kimura et al., 1979; Kimura, 19851. (However, i t is noteworthy that there is some question of the applicability of the findings of Kimura and colleagues to high speed gaits [see Vilensky, 19891.) CONCLUSIONS 1. The overall displacement patterns of the limb joints are similar during leading and trailing conditions. 2. Despite the similarity, there are notable differences in the values of some of the inflection points for the joints when the associated limb is leading or trailing. 3. In some cases the differences in joint values probably reflect only differing mechanical conditions, whereas in others, differences in muscle activation patterns are likely. 4. The facts that the monkeys were not always similar among themselves (especially the vervets compared to the squirrel monkey) and that a s a group they deviated from other species indicate that different animals may use slightly different biomechanical strategies during galloping. 5. It is possible that the asymmetry in the neural control programs associated with the leading and trailing limbs is a causal factor in the frequent lead changes commonly seen in animals. ACKNOWLEDGMENTS We are grateful to Mr. Gregory Duncan for assistance during filming, Ms. Diana Jackson for typing the manuscript, and Ms. Roberta Shadle for drawing some of the illustrations. Support for this study was provided by the Indiana University School of Medicine. REFERENCES Cohen, A.H. Relationship between forelimb coordination- and movement asymmetries during fast gaits, canter and gallop. BRAIN RESEARCH 164:352-356, 1979. Cohen, A.H.; Gans, C. Muscle activity in rat locomotion: Movement analysis and electromyography of the flexors and extensors of the elbow. JOURNAL OF MORPHOLOGY 146:177-195, 1975. Dagg, A.I. RUNNING, WALKING AND JUMPING. New York, Crane, Russak and Company, 1977. Hildebrand, M. Analysis of asymmetrical gaits. JOURNAL OF MAMMALOGY 58: 131-156, 1977. Howell, A.B. SPEED IN ANIMALS. Chicago, University of Chicago Press, 1944. Jenkins, F.A. Jr. Limb posture and locomotion in t h e Virginia opossum (Didelphis marsupialis) and i n other cursorial mam- Galloping in Monkeys I 239 mals. JOURNAL OF ZOOLOGY 165:303315, 1971. Kimura, T. Bipedal and quadrupedal walking of primates: Comparative dynamics. Pp. 81-104 in PRIMATE MORPHOPHYSIOLOGY, LOCOMOTOR ANALYSIS AND HUMAN BIPEDALISM. S. Kondo, ed. Tokyo, University of Tokyo Press, 1985. Kimura, T.; Okada, M.; Ishida, H. Kinesiological characteristics of primate walking: Its significance in human walking. Pp. 297-311 in ENVIRONMENT, BEHAVIOR AND MORPHOLOGY: DYNAMIC INTERACTIONS IN PRIMATES. M. Morbeck, H. Preuschoft, N. Gomberg, eds. New York, Fisher, 1979. Larson, S.G.; Stern, J.T. J r . The role of propulsive muscles of the shoulder during quadrupedalism in vervet monkeys (Cercopithecus aethiops): Implications for neural control of locomotion in primates. JOURNAL OF MOTOR BEHAVIOR 21:457472, 1989. Miller, S.; Van der Burg, J.;Van der Meche, F.G.A. Locomotion in the cat: Basic programmes of movement. BRAIN RESEARCH 91:239-253, 1975. Miller, S.; Van der Meche, F.G.A. Movements of the forelimbs of the cat during stepping on a treadmill. BRAIN RESEARCH 91~255-269,1975. Tokuriki, M. Electromyographic and jointmechanical studies in quadrupedal locomotion. 111. Gallop. JAPANESE JOURNAL OF VETERINARY SCIENCE 36:121-132, 1974. Vilensky, J.A. Primate quadrupedalism: How and why does it differ from that of typical quadrupeds? BRAIN BEHAVIOR AND EVOLUTION 34:357-364, 1989. Vilensky, J.A.; Gankiewicz, E.; Townsend, D.W. Effects of size on vervet (Cercopithecus aethiops) gait parameters: A crosssectional approach. AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 76:463-480, 1988. Vilensky, J.A.; Gankiewicz, E.; Townsend, D.W. Effects of size on vervet (Cercopithecus aethiops) gait parameters: A longitudinal approach. AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 81: 429-439,1990. Vilensky, J.A.; Patrick, M.C. Gait characteristics of two squirrel monkeys, with emphasis on relationships with speed and neural control. AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 68: 429-444,1985. Wetzel, M.C.; Anderson, R.C.; Brady, T.H. Jr; Norgren, K.S. Kinematics of treadmill galloping by cats. 111. Coordination during gait conversions and implications for neural control. BEHAVIORAL BIOLOGY 21: 107-127, 1977. Wetzel, M.C.; Stuart, D.G. Ensemble characteristics of cat locomotion and its neural control. PROGRESS IN NEUROBIOLOGY 7:l-98, 1976.