Energy expenditure of bipedal walking is higher than that of quadrupedal walking in Japanese macaques.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 131:33–37 (2006) Energy Expenditure of Bipedal Walking Is Higher Than That of Quadrupedal Walking in Japanese Macaques M. Nakatsukasa,1* E. Hirasaki,2 and N. Ogihara1 1 2 Laboratory of Physical Anthropology, Kyoto University, Kyoto 606-8502, Japan Laboratory of Biological Anthropology, Osaka University, Osaka 565-0871, Japan KEY WORDS locomotor energetics; bipedalism; quadrupedalism, Macaca fuscata; presteady state; steady state ABSTRACT The authors previously compared energetic costs of bipedal and quadrupedal walking in bipedally trained macaques used for traditional Japanese monkey performances (Nakatsukasa et al.  Am. J. Phys. Anthropol. 124:248–256). These macaques used inverted pendulum mechanics during bipedal walking, which resulted in an efﬁcient exchange of potential and kinetic energy. Nonetheless, energy expenditure during bipedal walking was signiﬁcantly higher than that of quadrupedal walking. In Nakatsukasa et al. ( Am. J. Phys. Anthropol. 124:248–256), locomotor costs were measured before subjects reached a steady state due to technical limitations. The present investigation reports sequential changes of energy consumption during 15 min of walking in two trained macaques, using carbon dioxide production as a proxy of energy consumption, as in Nakatsukasa et al. ( Am. J. Phys. Anthropol. 124:248–256). Although a limited number of sessions were conducted, carbon dioxide production was consistently greater during bipedal walking, with the exception of some irregularity during the ﬁrst minute. Carbon dioxide production gradually decreased after 1 min, and both subjects reached a steady state within 10 min. Energy expenditure during bipedalism relative to quadrupedalism differed between the two subjects. It was considerably higher (140% of the quadrupedal walking cost) in one subject who walked with more bent-knee, bent-hip gaits. This high cost strongly suggests that ordinary macaques, who adopt further bent-knee, bent-hip gaits, consume a far greater magnitude of energy during bipedal walking. Am J Phys Anthropol 131:33–37, 2006. V 2006 Wiley-Liss, Inc. A previous study by the authors revealed that the energetic cost of bipedal walking was signiﬁcantly higher (20–30%) than that of quadrupedal walking in two Japanese macaques (Macaca fuscata) (Nakatsukasa et al., 2004). This result raised concerns about generalizations drawn from the ﬁndings of Taylor and Rowntree (1973), who revealed that the energetic costs of bipedal and quadrupedal walking were identical in two chimpanzees and two capuchin monkeys. Unfortunately, bipedal locomotor energetic data in nonhuman primates are still limited to these six individuals of three species. Nakatsukasa et al. (2004) used Japanese macaques engaging in traditional bipedal monkey performances (hereafter, performing macaques). These macaques use inverted pendulum mechanics during bipedal walking, which results in an efﬁcient exchange of potential and kinetic energy (Hirasaki et al., 2004; Ogihara et al., 2005). Thus, the observed bipedal locomotor costs should underestimate those of ordinary Japanese macaques. However, locomotor energetic data of ‘‘untrained’’ Japanese macaques have not been obtained because they are not suitable for use in energetic experiments. There were a few technical problems in the previous study. A major problem concerned the relatively short experimental duration (2 min on average). This was due to the difﬁculty of conducting a sufﬁcient number of experiments with long duration and changing conditions (velocity, mode of locomotion). It is generally assumed that at least several minutes are required for an animal to reach steady state from the resting state. Energetic costs during a presteady state are higher than those during a steady state (Taylor et al., 1982), and do not provide proper measures for interspeciﬁc comparisons. Nakatsukasa et al. (2004) argued that this did not hamper an evaluation of the relative costs of bipedalism to quadrupedalism in the same individuals, because the duration of the experimental session did not differ signiﬁcantly between these locomotor modes. However, the duration of the presteady state in Japanese macaques and the extent to which presteady-state costs are higher than those of the steady state remain unknown. This information is important for a comparison of the data in Nakatsukasa et al. (2004) with other data obtained by the standard method (e.g., Taylor et al., 1982). We recently succeeded in collecting locomotor energetic data for a longer duration (15 min) in two performing macaques, although the number of sessions was limited. Here, we tested our previous conclusions by investigating sequential changes of energy expenditure during 15-min experimental sessions. In addition, we evaluated the extent to which differences of bipedal walking C 2006 V WILEY-LISS, INC. C Grant sponsor: Japanese Society for the Promotion of Science; Grant sponsor: Biodiversity Research for the 21st Century COE (A14); Grant number: Biodiversity Research for the 21st Century COE (A14), Grant-in-Aid 16370104. *Correspondence to: Masato Nakatsukasa, Laboratory of Physical Anthropology, Graduate School of Science, Kyoto University, Sakyo, Kyoto 606-8502, Japan. E-mail: firstname.lastname@example.org Received 1 July 2005; accepted 9 November 2005. DOI 10.1002/ajpa.20403 Published online 16 February 2006 in Wiley InterScience (www.interscience.wiley.com). 0.473 0.436 0.387 0.432 0.354 0.319 0.320 0.331 0.325 0.294 0.300 0.313 0.302 0.297 0.261 0.284 0.340 0.297 4 3 2 1 Standard error calculated from LSR slope of CO2 concentration increase rates (in italics). Mean of minute averages of CO2 production rates for 3 min after outset of exercise. In parentheses, percent value to steady–state value (bottom row). Mean of minute averages of CO2 production rates for 3–6 min after outset of exercise. In parentheses, percent value to steady–state value (bottom row). Mean of minute averages of CO2 production rates for 10–15 min after outset of exercise (steady–state value). 0.011 0.007 0.005 (118.0) 0.007 0.010 0.011 (106.0) 0.006 0.013 0.013 0.006 0.008 0.012 0.013 0.010 0.010 0.420 0.389 0.382 0.397 0.365 0.344 0.359 0.356 0.362 0.352 0.348 0.343 0.325 0.367 0.353 0.309 0.323 0.336 0.013 0.014 0.008 (119.0) 0.013 0.018 0.010 (108.0) 0.007 0.018 0.008 0.007 0.011 0.007 0.011 0.014 0.008 0.555 0.507 0.503 0.521 0.487 0.464 0.473 0.475 0.426 0.436 0.450 0.462 0.443 0.407 0.431 0.453 0.463 0.439 0.010 0.009 0.007 (116.0) 0.007 0.010 0.007 (112.0) 0.007 0.013 0.007 0.008 0.008 0.013 0.007 0.010 0.015 0.404 0.444 0.428 0.426 0.405 0.411 0.410 0.409 0.405 0.397 0.381 0.392 0.354 0.380 0.379 0.354 0.360 0.366 0.014 0.019 0.016 (109.0) 0.007 0.005 0.009 (105.0) 0.020 0.005 0.015 0.011 0.011 0.009 0.010 0.015 0.021 0.428 0.446 0.412 0.429 0.417 0.411 0.410 0.413 0.398 0.383 0.403 0.391 0.389 0.395 0.381 0.399 0.394 0.392 0.081 0.020 0.010 (113.0) 0.028 0.060 0.091 (111.0) 0.015 0.026 0.012 0.050 0.046 0.017 0.023 0.012 0.017 0.413 0.360 0.363 0.379 0.375 0.382 0.357 0.372 0.356 0.319 0.324 0.352 0.376 0.322 0.322 0.330 0.322 0.335 0.021 0.012 0.012 (121.0) 0.013 0.010 0.019 (112.0) 0.010 0.006 0.013 0.012 0.023 0.016 0.033 0.015 0.014 0.465 0.400 0.380 0.415 0.382 0.372 0.388 0.381 0.366 0.331 0.318 0.324 0.347 0.330 0.365 0.356 0.309 0.341 0.455 0.506 0.495 0.485 0.485 0.495 0.439 0.473 0.458 0.457 0.407 0.427 0.407 0.420 0.447 0.417 0.398 0.418 0.482 0.489 0.433 0.468 0.495 0.448 0.444 0.462 0.442 0.433 0.411 0.431 0.419 0.417 0.413 0.416 0.409 0.415 0.048 0.039 0.039 (113.0) 0.014 0.035 0.018 (111.0) 0.028 0.030 0.021 0.078 0.040 0.021 0.057 0.015 0.040 0–1 min 1–2 min 2–3 min 0–3–min mean2 3–4 min 4–5 min 5–6 min 3–6–min mean3 6–7 min 7–8 min 8–9 min 9–10 min 10–11 min 11–12 min 12–13 min 13–14 min 14–15 min 10–15 min mean4 0.022 0.037 0.029 (116.0) 0.055 0.036 0.063 (113.0) 0.047 0.026 0.029 0.051 0.029 0.036 0.049 0.022 0.041 B, 6.1 kg Biped 1 E1 A, 10.0 kg Biped 2 E1 A, 10.0 kg Biped 1 E1 A, 8.2 kg Quad 2 E1 A, 8.2 kg Quad 1 E1 A, 8.2 kg Biped 2 E1 The subjects were two male Japanese macaques, maintained by the Suo Monkey Performance Association (Kumamoto Prefecture, Japan) and trained for bipedal performance from age 2 years. The Suo Monkey Performance is regarded as an intangible cultural asset by the local government, and is operated under municipal regulations with respect to animal welfare and treatment. Experiments were conducted with the collaboration of the Suo Monkey Performance Association, following the guidelines for animal experimentation of Kyoto University. Subject A had been used in previous experiments (subject 2 in Nakatsukasa et al., 2004). This macaque was an 8-year-old male as of May 2005. Experiments were conducted in September 2003 and May 2005, when his body mass was 8.2 kg and 10.0 kg, respectively. Subject B was a 5-year-old male, and possessed a body mass of 6.1 kg in May 2005. The experimental method used was similar to that outlined previously (Nakatsukasa et al., 2004). The size of the chamber was reduced by half to detect gas concentration more precisely (1.9 3 0.95 3 1.9 m). The capacity of the chamber was about 3,330 l, excluding the total volume of the subject, trainer, and experimental apparatus. The experimental duration was 15 min. If sessions were consecutive, at least 20 min of resting time were taken between sessions. The subjects walked on the treadmill set in the chamber at a velocity of 2.0 km/hr. They did not show apparent tiring after completing 15 min of walking at this velocity. The trainer accompanied the subject and breathed through a tube connected to the outside of the chamber. Carbon dioxide (CO2) concentrations within the chamber were recorded using an infrared gas analyzer (Model CGT-7000, Shimadzu Corp.) positioned in the chamber. The recording was started 30 sec after the chamber was closed, following a wait for the diffusion of the inner air. The data-collection intervals for experiments conducted in 2003 and 2005 were 10 sec and 5 sec, respectively. A least-squares linear regression (LSR) was calculated between CO2 concentration (ppm) and elapsed time for data recorded every minute. The LSR slope represented the average CO2 concentration increase for 1 min. The standard error of the LSR slope was also calculated. The slope was converted into a standardized CO2 production rate (ml/kg/sec) (Table 1). Although this value is not the same as the true energetic cost given by oxygen consumption, CO2 production can be used as a convenient measure of energy consumption under the assumption that the respirator quotient was constant during experiments (Nakatsukasa et al., 2004). It is known empirically that the respirator quotient is in the range of 0.8– 0.9 in animals fed with balanced diets. The bipedal kinematics of subject A were reported in Hirasaki et al. (2004) as the third subject among the performing macaque group. Kinematic data of subject B were taken in October 2004 by a common method using two CCD camera systems and digitizing software on a PC, as in Hirasaki et al. (2004). TABLE 1. Sequential changes of standardized CO2 production (ml/kg/sec) calculated for each minute during 15-min walking MATERIALS AND METHODS B, 6.1 kg Quad 1 E1 B, 6.1 kg Quad 2 E1 style affected energy expenditure in these Japanese macaques. Bipedal kinematics for one of the two performing macaques involved a more bent-knee, bent-hip style, with the trunk more inclined in comparison to the other experimental subject. The development of inverted pendulum mechanics during bipedal walking was at an incipient stage. 0.020 0.012 0.013 (145.0) 0.007 0.012 0.009 (111.0) 0.010 0.007 0.009 0.011 0.010 0.009 0.007 0.018 0.015 M. NAKATSUKASA ET AL. A, 8.2 kg Subject, body mass Trial Biped 1 E1 34 LOCOMOTOR ENERGETICS IN JAPANESE MACAQUE Fig. 1. Sequential changes of standardized CO2 production (ml/kg/hr) during 15-min walking exercise in subject A in 2003: two trials for bipedal and quadrupedal walking (thin broken and dotted lines), and their averages (thick solid lines). CO2 production was calculated for each minute from outset of exercise. 35 Fig. 2. Sequential changes of standardized CO2 production (ml/kg/hr) during 15-min walking exercise in subject A in 2005: two trials for bipedal walking (thin broken and dotted lines), and their average (thick solid line). CO2 production was calculated for each minute from outset of exercise. RESULTS Rates of CO2 production and their standard errors were measured in two subjects under different conditions, and the results are summarized in Table 1 and shown in Figures 1–3. Figure 1 represents results obtained from subject A when he was 6.3 years of age. Two sessions were recorded for both bipedal and quadrupedal walks. The average of the two sessions in each locomotor mode is also shown. Values after the ﬁrst minute (left endpoint) are rather irregular in comparison to later values. They are rather high in quadrupedal sessions, and low in bipedal sessions. However, CO2 production rates exhibit gradual decreasing tendencies after the ﬁrst minute, although some ﬂuctuations are observed. The values for the bipedal walk are always higher than corresponding values of the quadrupedal walk. The decreasing tendency is absent 8 min after the outset of the experiment, indicating that the subject reached a steady state. Combining the two sessions, the grand mean of the last 5 min of data is 0.417 (ml/kg/sec) for the bipedal walk and 0.338 for the quadrupedal walk, with the former being 23% higher than the latter. Figure 2 represents CO2 production rates in subject A at age 8 years. Only bipedal walk data were recorded. The observed features are the same as those shown in Figure 1. In this case, the subject might have reached a steady state somewhat later, at 10 min after the outset of the sessions. The grand mean of the last 5 min of data is 0.379 (ml/kg/sec). Owing to the increased body mass (þ1.8 kg) from 2003, the locomotor cost of a bipedal walk per body mass is reduced by 10% (Nakatsukasa et al., 2004). Figure 3 represents the CO2 production rate of subject B during bipedal and quadrupedal walks. Two sessions were recorded for the quadrupedal walk, and one session for the bipedal walk. Average values are shown for the quadrupedal sessions. Although ﬂuctuations exist during later phases of all sessions, this subject probably reached a steady state around 8 min after the onset of the experiment. CO2 production values for the bipedal walk are always higher than corresponding quadrupedal values. The average of the last 5 min of data is 0.439 (ml/ kg/sec) in the bipedal walk, and 0.317 in the quadrupedal walk. Thus, the bipedal cost at steady state is about Fig. 3. Sequential changes of standardized CO2 production (ml/kg/hr) during 15-min walking exercise in subject B: one trial for bipedal walking, and two for quadrupedal walking (thin solid, broken, and dotted lines). Thick solid line represents average of quadrupedal data. CO2 production was calculated for each minute from outset of exercise. 40% higher than that recorded during a quadrupedal walk. DISCUSSION The results of this study support the conclusion in our previous study (Nakatsukasa et al., 2004). Although CO2 production rates ﬂuctuate irregularly during the ﬁrst minute of experiments in some cases (Fig. 1), they decrease in both bipedal and quadrupedal sessions in a parallel manner. The irregularity of CO2 production soon after exercise may be inﬂuenced by temporary storing of CO2 in the blood and muscles. Although the experimental sessions in Nakatsukasa et al. (2004) were generally short, the present results indicate that energy expenditure during bipedal walking relative to quadrupedal walking (B/Q ratio) in the ﬁrst 2 min is more or less similar to that recorded in later stages, though it is not coincident. Clearly, bipedal walking at a steady state in performing macaques needs more energy than quadrupedal walking at a steady state. Table 1 compares the presteady state CO2 production rates with those at steady state (¼ mean of 10–15 min of data). Average CO2 production rates in the ﬁrst 3 min vary from 109–145% of those recorded at steady state, 36 M. NAKATSUKASA ET AL. with a grand mean of 119%. Many of the experimental sessions in Nakatsukasa et al. (2004) fall into this range. Thus, their values are likely to be higher than those at steady state by 20% on average. Average CO2 production rates during 3–6 min after the outset of exercise are constantly higher than rates recorded at steady state during either bipedal or quadrupedal walking. They are 105–113% (110% on average) of steady-state values. In all sessions, subjects reached a steady state within 10 min from the outset of exercise. However, if practical limitations hamper the continuation of such a long experiment, a compromise may be to use data covering a period of 3–6 min. Although they overestimate energy consumption at steady state, B/Q ratios during this period are similar to those at steady state (Table 1). The B/Q ratio at steady state in subject A is 124%. This ratio is consistent with the ﬁndings of previous experiments (Table 3 of Nakatsukasa et al., 2004), when the subject was 4–5 years of age. Thus, the relative energetic cost of bipedalism to quadrupedalism did not change signiﬁcantly following 3 years’ growth in this individual. This is concordant with the fact that kinematic features of this individual did not change very much between 2001 and 2004 (E. Hirasaki, unpublished data). The B/Q ratio in subject B at steady state (138%) is much higher than that recorded for subject A (and subject 1 in Nakatsukasa et al., 2004). Since data for the bipedal walk were obtained from only one session, and the two quadrupedal sessions displayed large ﬂuctuations, care must be taken not to rely too heavily on these data. Nonetheless, this individual apparently consumed a relatively greater amount of energy during bipedalism in comparison to other performing macaques. Figure 4 compares stick diagrams of the bipedal walk (at ca. 3.5 km/hr) in performing and ordinary macaques. These stick diagrams were recorded using the method described in Hirasaki et al. (2004). Subject B shows a subtle upward convex trajectory of the hip joint during the support phase (inverted pendulum-like mechanics). His trunk is more inclined, and the hip and knee joints are less extended, compared to other performing macaques. Thus, he is not a good biped among performing macaques. However, his hip and knee joints are much more extended in comparison to ordinary macaques, and he is still a better biped compared to ordinary macaques. It is reasonable to use these bipedal locomotor characteristics to explain the relatively high energetic consumption of bipedal walking. Data for these stick diagrams were obtained from faster bipedal walks, conducted at a velocity of approximately 3.5 km/hr. It should also be noted that these bipedal walking characteristics, except for hip-joint motions, in both performing and ordinary macaques are not affected signiﬁcantly by walking velocity (Hirasaki et al., 2004). Although hipjoint motion is affected by walking speed, differences between performing and ordinary subjects are maintained, regardless of walking speed. According to the biomechanical theoretical analysis of Ishida (1991), ‘‘untrained’’ macaques consume 50% more energy during a bipedal walk compared to well-trained (¼ performing) macaques, while ‘‘moderately trained’’ macaques consume 10% more energy in a similar comparison. Subject B consumes about 10% extra energy during a bipedal walk when compared to subject A Fig. 4. Stick diagrams of bipedal walking at velocity of approximately 3.5 km/hr. a–c: Performing macaques. a, b: Subjects 1 and 2 in Nakatsukasa et al. (2004). c: Subject B. d, e: Ordinary experimental macaques. Note that subject B walks with more inclined trunk and less extended hip and knee joints compared to other performing macaques (a, b). However, hip and knee joints are still more extended compared to ordinary macaques. (138% vs. 124% in B/Q ratio), given that their quadrupedal walking requires a constant amount of energy. This result is close to the prediction of Ishida (1991). Although the above calculation is based on a small number of experiments and data improvement is necessary, it strongly suggests the rather high energy consumption of bipedal walking in ordinary Japanese macaques, who may be categorized as ‘‘untrained’’ or ‘‘bad’’ bipeds (Ogihara et al., 2005). Cercopithecines generally share similar postcranial skeletal features in relation to ancestral semiterrestrial habits (Schultz, 1970; Strasser, 1988; Strasser and Delson, 1987; Harrison, 1989; Nakatsukasa, 1994). If the relative cost of a bipedal walk is so expensive in ordinary Japanese macaques, it is unlikely that energy expenditures in bipedal and quadrupedal walking are the same in other cercopithecine monkeys (and very likely colobine monkeys). Locomotor energetic experiments on primates are difﬁcult. However, to test the conclusion of Taylor and Rowntree (1973), comparable data on platyrrhines should be collected. Our study conﬁrmed that variation of gait features such as trunk angle, hindlimb joint angles, and trajectory of the center of mass signiﬁcantly affect the energetic cost of bipedalism in nonhuman primates. This result was not surprising, as it had already been predicted by inverse dynamic simulation (Yamazaki, 1985; Ishida, 1991). Nonetheless, experimental data are important to reproduce the predictions of inverse dynamic simulation models. Carey and Crompton (2005), for example, demonstrated that a usage of emphasized bent-hip, bentknee bipedalism (as in Fig. 4d,e) considerably overloaded locomotor physiology in modern humans, compared to LOCOMOTOR ENERGETICS IN JAPANESE MACAQUE fully upright bipedalism. However, even comparatively ‘‘minor’’ differences of bipedal gait characteristics (Fig. 4a–c) introduce signiﬁcant inﬂuence on the energetic cost of bipedal locomotion. Improvement of locomotor efﬁciency allows an animal to relocate the saved energy for other purposes, such as growth or reproduction, or to have a wider day range. This result, therefore, suggests that if early hominids had traveled daily over long distances, their gait characteristics might be an acute target of natural selection. However, identifying the most proximate target of selection on the locomotor musculoskeletal system is not straightforward. Stern (1999) proposed that a reduction of peak vertical force and an increase of speed were beneﬁcial in Australopithecus afarensis, even if bent-hip, bent-knee bipedalism sacriﬁced locomotor energetic efﬁciency. Susman et al. (1984) argued that the less developed sacro-iliac stabilization mechanism in A. afarensis could be related to the low peak vertical force in benthip, bent-knee bipedalism (but see opposing argument in Carey and Crompton, 2005). Steudel-Numbers and Tilkens (2004) presented a similar case, where experiments clearly indicated that an elongation of the lower limb signiﬁcantly contributed to a reduction of bipedal locomotor cost. However, the lower limb of early australopithecines (A. afarensis and A. africanus) stayed comparatively short for at least half a million years. Here also, another target (or other targets) of selection must be presumed, such as lateral balance control during the single stance phase and/or climbing capability (SteudelNumbers and Tilkens, 2004). Locomotor energetics is a very important component for understanding the paleobiology of early hominids and the transition to habitual bipedality in protohominids. However, it will be most useful when it is holistically synthesized with information on hominids’ paleoecology and their skeletal functional morphology. CONCLUSIONS We conﬁrmed that highly trained bipedal performing macaques consumed more energy during bipedal locomotion compared to quadrupedal walking through 15-min exercise experiments. The subjects reached steady state within 10 min in both bipedal and quadrupedal sessions. The presteady extra cost was about 20% of the steadystate cost rates in the ﬁrst 3 min, and about 10% during 3–6 min after the outset of exercise. The energetic cost of bipedal walking relative to quadrupedal walking was 124% in one subject and 138% in the other. The relatively high cost in the latter subject was related to his bipedal gait characteristics such as a more inclined trunk, and less extended knee and hip joints. However, this subject was an appreciably ‘‘better’’ biped (less benthip, bent-knee) compared to untrained Japanese macaques. Therefore, the relative bipedal energetic cost should be rather high in untrained macaques. The same result probably holds true for Old World monkeys in general. 37 ACKNOWLEDGMENTS We are very grateful to the trainers of the Suo Monkey Performance Association for their enormous patience in carrying out the repetitive experiments required of this investigation. This study was made possible by the very generous support received from the Suo Monkey Performance Association. LITERATURE CITED Carey TS, Crompton RH. 2005. The metabolic costs of ‘‘bent-hip, bent-knee’’ walking in humans. J Hum Evol 48:25–44. Harrison T. 1989. New postcranial remains of Victoriapithecus from the Middle Miocene of Kenya. J Hum Evol 18:3–54. Hirasaki E, Ogihara N, Hamada Y, Kumakura H, Nakatsukasa M. 2004. Do highly trained monkeys walk like humans? A kinematic study of bipedal locomotion in bipedally trained Japanese macaques. J Hum Evol 46:739–750. Ishida H. 1991. A strategy for long distance walking in the earliest hominids: effect of posture on energy expenditure during bipedal walking. In: Coppens Y, Senut B, editors. Origine(s) de la bipédie chez les hominidés. Paris: CNRS. p 9–15. Nakatsukasa M. 1994. Morphology of the humerus and femur in African mangabeys and guenons: functional adaptation and implications for the evolution of positional behavior. Afr Stud Monogr [Suppl] 21:1–61. Nakatsukasa M, Ogihara N, Hamada Y, Goto Y, Yamada M, Hirakawa T, Hirasaki E. 2004. Energetic costs of bipedal and quadrupedal walking in Japanese macaques. Am J Phys Anthropol 124:248–256. Ogihara N, Usui H, Hirasaki E, Hamada Y, Nakatsukasa M. 2005. Kinematic analysis of bipedal locomotion of a Japanese macaque that lost its forearms due to congenital malformation. Primates 46:11–19. Schultz AH. 1970. The comparative uniformity of the Cercopithecidae. In: Napier JR, Napier PR, editors. Old World monkeys: evolution, systematics and behavior. London: Academic Press. p 39–51. Stern JT. 1999. The cost of bent-knee, bent-hip bipedal gait. J Hum Evol 36:567–570. Steudel-Numbers KL, Tilkens MJ. 2004. The effect of lower limb length on the energetic cost of locomotion: implications for fossil hominins. J Hum Evol 47:95–109. Strasser E. 1988. Pedal evidence for the origin and diversiﬁcation of cercopithecid clades. J Hum Evol 17:225–245. Strasser E, Delson E. 1987. Cladistic analysis of cercopithecid relationships. J Hum Evol 16:81–99. Susman RL, Stern JT, Jungers WL. 1984. Arboreality and bipedality in the Hadar hominids. Folia Primatol (Basel) 43: 113–156. Taylor CR, Rowntree VJ. 1973. Running on two or on four legs: which consumes more energy? Science 179:186–187. Taylor CR, Heglund NC, Maloiy GMO. 1982. Energetics and mechanics of terrestrial locomotion. J Exp Biol 97:1–21. Yamazaki N. 1985. Primate bipedal walking: computer simulation. In: Kondo S, editor. Primate morphophysiology, locomotor analyses and human bipedalism. Tokyo: University of Tokyo Press. p 105–130.