Traditionally, the functional interpretation of fossils that document our earliest locomotor evolution would be based on morphological comparisons to modern humans and a selection on nonhuman primates, most commonly, our nearest living relatives, the great apes. While once it was considered sufficient to simply label a trait as ‘humanlike’ or ‘apelike,’ current standards require empirical support for the functional interpretation of features. Research documenting the biomechanics of human locomotion is readily available, and for comparisons, bipedal locomotion has been explored in several nonhuman primates including not only the great apes but also lesser apes and a variety of Old and New World monkeys. Which taxon is examined and the type of information gathered naturally depend on the questions being asked, and the scope and diversity of this large body of research on nonhuman primate locomotion has been the topic of several reviews including Fleagle (1979), Jouffroy (1989), Schmitt (2003), D'Août et al. (2014), and Druelle and Berillon (2014).

Studies on the locomotor mechanics of chimpanzees and bonobos have figured prominently in comparative studies of bipedal locomotion, and as Pontzer et al. (2014: p.79) have argued, this is not so much because Pan is the perfect model for understanding early hominin locomotor evolution “but because it enables us to test biomechanical models of locomotor performance in a large-bodied semi-arboreal primate.” Much is known about the forces that their limbs exchange with the ground during both bipedalism and knuckle-walking (e.g., Kimura et al., 1977, 1979; Kimura, 1985, 1992; Reynolds, 1985a,b; Demes et al., 1994, 2015; Li et al., 1996; Vereecke et al., 2003; Sockol et al., 2007; Pontzer et al., 2014; O'Neill et al., 2022) and about the kinematic characteristics of both over-ground locomotion and climbing (e.g., Jenkins, 1972; Reynolds, 1987; Aerts et al., 2000; D'Août et al., 2002, 2004; Isler, 2002, 2005; Pontzer et al., 2014; O'Neill et al., 2015, 2018; Holowka et al., 2017a,b; Thompson et al., 2018; Fannin et al., 2023; Druelle et al., 2024b). However, there are much less electromyographic (EMG) data on hind limb muscle recruitment patterns during locomotion in chimpanzees and none on bonobos. Muscles are the main force producers supporting the body and moving limb segments during locomotion, and many functional interpretations of morphology, particularly when it is applied to fossils, are related to muscular anatomy and inferred muscle function. Examples include features related to muscle leverage such as the orientation of the iliac blades of a hominin pelvis and the role of the lesser gluteals (e.g., Lovejoy et al., 1973; Stern and Susman, 1981, 1983) or differences in ischium length and hamstring leverage for walking or climbing (e.g., Robinson, 1972; Lovejoy et al., 1973; McHenry, 1975; Pontzer et al., 2009; Foster et al., 2013; Lewton and Scott, 2017; Kozma et al., 2018). In addition to muscle leverage, features related to the size and line of action of a muscle have been used to functionally interpret fossils such as whether grooves created by the tendon of the obturator externus on the posterior aspect of the hominin femoral neck (Day, 1969; Lovejoy, 1975; Lovejoy et al., 1973; Stern and Larson, 1993; Pickford et al., 2002) or for the iliopsoas in the pelvic brim (Lovejoy, 1988) are related to hip posture, or what a large peroneal trochlea on a calcaneus and relatively broad and deep grooves for the peroneal muscles on fibulae (Stern and Susman, 1983) indicate about ankle function, or if the existence of a large popliteal groove on the distal femur is related to a hyperpronated gait in Australopithecus sediba (DeSilva et al., 2013). Basic knowledge of the contributions of individual muscles to locomotor behaviors is therefore essential to validate such inferences. However, the activity patterns of only a handful of chimpanzee hip and thigh muscles have been examined (Tuttle et al., 1978, 1979; Fleagle et al., 1981; Stern and Susman, 1981; Jungers et al., 1983; Ishida et al., 1985; Kumakura, 1989; Stern and Larson, 1993; Larson and Stern, 2009), and for those few that have been studied, the data that exist are often limited to raw EMG for single strides (e.g., Jungers et al., 1993) with only qualitative indications of relative amplitude or variability in recruitment.

The aim of many of those studies that have explored chimpanzee hind limb muscle activity patterns has been to look for patterns of muscle activations indicative of a functional preadaptive behavior for hominin bipedalism. Stern and Susman (1981) compared patterns of recruitment of the gluteal musculature during climbing, bipedalism, and quadrupedalism in chimpanzees and orangutans and found marked similarities between climbing and bipedalism and dissimilarities of both to quadrupedalism for chimpanzees, although not for orangutans. On the other hand, Kumakura (1989) observed that the chimpanzee long head of biceps femoris displays comparable patterns of recruitment during climbing, bipedalism, and quadrupedalism. However, in regard to comparisons of muscle activity patterns in chimpanzees to those of humans, other than the parallels in gluteus medius use by chimpanzees during bipedalism and climbing to human lesser gluteal recruitment (Stern and Susman, 1981, see below), the activity patterns of the chimpanzee gluteus superficialis or long head of biceps femoris during both bipedal walking and climbing are unlike those of the human gluteus maximus or biceps femoris during human bipedalism (Stern and Susman, 1981; Kumakura, 1989; Knutson and Sonderberg, 1995).

While these studies examined only a small number of muscles, a recent paper has explored the question of a possible preadaptive behavior with a more complete list of chimpanzee hind limb muscles by applying muscle synergy analysis to the patterns of recruitment (Goto et al., 2024). Rather than the exploration of individual muscle activations, muscle synergy analysis attempts to explore patterns of neural control by identifying subsets of muscle groups, known as synergies, that simplifies movement production by the modular organization of groups of coactivating muscles (for more information on muscle synergy analysis, see Grillner, 1985; Ivanenko et al., 2004). Goto et al. (2024) reported that synergies observed during the stance phase of chimpanzee bipedal walking were similar to those of vertical climbing, while synergies during swing phase were more similar to those during knuckle-walking, although there was some mosaicism in each comparison.

In the absence of information on patterns of muscle activations, modeling approaches offer a means of gaining some insights into the contributions of muscle force during chimpanzee locomotion. For example, Pontzer et al. (2009) estimated locomotor costs using a ‘force production’ approach, which predicts costs from the rate at which muscle force is generated to support body weight, to compare the metabolic cost of walking in humans, chimpanzees, and early hominins. Calculation of the volume of muscle activated during locomotion is based on the magnitude of ground reaction force impulses, the length of muscle fascicles, and the ratio of the anatomical moment arm of the muscles to the load arm of the ground reaction force vector. While a useful way to explore how even minor changes in limb length, effective mechanical advantage, or muscle fascicle length might decrease the cost of locomotion in an early hominin biped, the results are difficult to apply to the interpretation of fossil material, which are based on a higher level of anatomical detail, including individual muscle attachments. The calculations are based on functional groupings of muscles, which presumes knowledge of the functional roles of all chimpanzee limb muscles. However, the fact that a muscle is capable of performing a particular action at a joint does not necessarily mean it is used to do so. For example, the latissimus dorsi and teres major are humeral retractors that are used to pull the body past the supporting forelimb in most mammalian quadrupeds but are not used to do so in knuckle-walking chimpanzees (Larson and Stern, 1987).

Inverse dynamics is another modeling approach that can indicate the contributions of muscular force. Using kinematics, ground reaction forces, and limb segment inertial parameters to calculate net forces, moments, work, and power at a joint, O'Neill et al. (2022) used this approach to compare the mechanics of chimpanzee bipedal walking to that of human walking in order to evaluate hypotheses relating derived human features to improved effectiveness of human bipedal gait. However, as with the study by Pontzer et al. (2009), trying to apply their observations to the interpretation of fossil material would be facilitated by knowing which muscles are active at specific times during a chimpanzee bipedal stride that contribute to observed joint moments.

The studies by Pontzer et al. (2009), O'Neill et al. (2022), and Goto et al. (2024) all used broad scales of analysis appropriate to the questions being asked and were not designed to reveal the functional roles of individual chimpanzee hind limb muscles. However, knowing those details can be important. For example, for many years researchers incorrectly interpreted the lesser gluteals of apes to be hip extensors, leading to erroneous interpretations of pelvic anatomy in early hominins (see summary in Stern and Susman, 1981). An EMG study on lesser gluteal activity in chimpanzees, orangutans, and gibbons, however, revealed that these muscles play the same role of lateral balance control during ape bipedal walking as they do in humans (Stern and Susman, 1981). In apes, they do so by virtue of their action as medial rotators of a flexed hip, whereas in humans, they do so by virtue of their action as abductors of an extended hip, putting a different light on the interpretation of iliac blade orientation of early hominin pelves (e.g., Schmidt, 1983; Stern and Susman, 1983, 1991). Understanding how differences in muscle function may impact morphology in fossils still requires detailed knowledge of the locomotor contribution of individual muscles.

More recent chimpanzee musculoskeletal models such as that of O'Neill et al. (2013) incorporate a higher level of anatomical detail, including representations of all of the major individual muscles of the hind limb, and facilitate more in-depth explorations of chimpanzee locomotor mechanics. Approaches such as static optimization modeling using inverse dynamics or forward simulation will benefit from detailed EMG data since they can be used to validate predictions or be used to drive simulation studies.

The current study addresses this need by presenting the EMG-based activity profiles for 27 chimpanzee hind limb muscles during knuckle-walking, bipedalism, and vertical climbing, with separate profiles for individual components of multipart muscles such as the quadriceps femoris or gastrocnemius for a total of 37 muscle activity profiles. This list includes nearly all of the muscles of the hip, thigh, and leg, except the short rotators of the hip (e.g., the superior and inferior gemellus), but only two intrinsic foot muscles (the adductor and abductor hallucis). The three locomotor behaviors are the ones discussed most frequently in regard to the interpretation of hominin fossil material, although they of course do not represent a chimpanzee's entire behavioral repertoire (see for example, Hunt, 1992, or Sarringhaus et al., 2014). This study makes no a priori assumptions about functional groupings of muscles and presents simple side-by-side comparisons of the recruitment of individual muscles during different behaviors, facilitating the determination of which hind limb muscles are major contributors to particular locomotor behaviors.

Most previous studies documenting activity patterns of chimpanzee hind limb muscles are based on one or two subjects, which is typical for lab-based research on chimpanzees as well as other nonhuman primates due to limited subject availability, space, and resources. Some information is better than none, but questions about how representative a particular activity pattern is or whether there is variability in recruitment cannot be addressed. The current study takes advantage of the long history of the Stony Brook Primate Locomotion Lab to compile EMG data from eight chimpanzee subjects. Nearly complete datasets have been collected for the two most recent chimpanzee subjects, which are supplemented with more limited samples of muscle recruitment data from six previous subjects. For many muscles, therefore, it will be possible to explore variability in activity among individuals. Just as there are morphological and kinematic differences between individuals, it is expected that there will be individual variation in EMG profiles as well. However, it is reasonable to expect that a critical contribution of muscular force to a locomotor behavior will display the least intersubject variability, whereas high variability might indicate a greater potential for idiosyncrasy in the performance of the behavior. The presentation of activity patterns during knuckle-walking, climbing, and bipedal walking will indicate which of these behaviors make the greatest demands on particular muscles, which can be helpful in determining a muscle's functional role. Finally, the muscle activity patterns of the eight subjects can be compared to explore for the possible effects on training in regard to one particular behavior—bipedal walking. Although bipedalism is not a frequent chimpanzee behavior in the wild and is considered to be a facultative form of locomotion for them, it continues to figure prominently in theories of hominin evolution (e.g., Drummond-Clarke et al., 2022). The two most recent chimpanzee subjects were part of a project to document the kinematics and kinetics of chimpanzee bipedalism, and the greater frequency with which they were encouraged to walk bipedally resulted in them being more practiced at walking on their hind limbs than any of the previous subjects. In general terms, the acquisition of a skilled behavior involves the suppression of superfluous muscle activations in favor of a coordinated recruitment pattern that effectively accomplishes the task (Basmajian and DeLuca, 1985; Thoroughman and Shadmehr, 1999; Carson and Riek, 2001). If there is a detectable training effect, the two most recent chimpanzee subjects that were more practiced at bipedal walking than previous subjects would be expected to display some reductions in muscle activations and/or decreases in amplitude of recruitment.