Interspecific differences in dorsiflexors
The galago has unique PALs for the dorsiflexors. The galago had shorter PALs and lower MAs of the TA and EHL than those of the loris and lemur. A morphological interpretation is that drastically extended tarsal bones alter the lines of actions of the dorsiflexors. The insertion of the TA lies in the navicular bone, and the EHL tendon reaches the big toe via a retinaculum near the insertion of the TA (Stevens et al. 1982). Therefore, the elongated tarsal bones displace the insertion of the TA and the intermediate point of the EHL toward the distal part of the foot, resulting in the lines of action forming sharper angles relative to the plane of the foot (Fig. 6). Because PAL is defined as the shortest distance between the line of action and the joint axis (Zajac, 1992; Gebo, 1993; Thorpe et al. 1999; Payne et al. 2006), the PALs are affected by the angle between the lines of actions of the dorsiflexors and the plane of the foot. In this case, the foot morphology of the galago results in the short PALs of the dorsiflexor, and MAs are relatively low compared with the others.
Figure 6. The line of action of the TA. In the galago, the elongated tarsal bones displace the insertion of the TA to the distal part of the foot. The acute angle between the line and long axis of the foot results in the short PAL of the TA. The star symbol represents the angle.
Download figure to PowerPoint
When leaping, the galago extends its hind limb joints more than the lemur does (Günther et al. 1991). This is especially true for Galago senegalensis, which is an extremely specialised leaper with small body size (Crompton et al. 1987; Günther et al. 1991). The angular excursions of the hind limb joints are the most pronounced in the small galago during the final stage of take-off (Günther et al. 1991). Although Galago garnettii has a larger body size than the small galago, the angular excursions of the joints are larger and faster than those of the lemur (Günther et al. 1991). It is likely that the difference in the angular velocity also partly reflects an interspecific difference in the distal segment weight of the hind limb in each species. However, for the galago to prepare to land securely on substrates, the large ankle joint extension would need to be recovered quickly or dorsiflexed immediately after take-off. In that case, the relatively low MAs of the dorsiflexors would be advantageous in their locomotion.
Interspecific differences in planterflexors
A general trend was found in the interspecific differences among PALs of most plantarflexors in the studied groups. In descending order of PAL, plantarflexors (i.e. SL, LG, MG, and PT) values were distributed as follows: galago, lemur, and loris. However, the trend was reversed for MAs. This result suggests that the LALs of these species also vary, as do the PALs, and that the MAs are affected not only by PALs but also by foot morphology in these species.
The loris moves with slow quadrupedalism in the canopy and uses horizontal supports (Grand, 1967; Crompton et al. 1987). Gebo (1987) reported that the loris uses bridging and suspensory behaviour to fill the gaps between supports much more frequently than does the lemur in arboreal environments. In addition, lemurs do not using cantilevering, in which a primate orients outward from a vertical support and holds a horizontal position. In contrast, the loris performs this positional behaviour with its hind limbs while foraging in an arboreal environment (Gebo, 1987). In this posture, the centre of mass is located away from the base of support, where the feet are firmly gripping twigs. In the horizontal posture, the moment of external force that the ankle joint must bear is proportional to the distance between the feet and the centre of mass; therefore, the muscle should produce a high joint moment to stabilise this posture. Thus, the loris would have high MAs for plantarflexion to respond to this mechanical demand. In addition to stabilising the joint, the loris must also grasp twigs firmly in this positional behaviour. Interestingly, the flexors of digits, particularly the FDF, yielded significantly higher MAs in the loris than in the other species.
The MAs of most plantar flexors were higher in the lemur than in the galago. The lemur and galago have similar locomotor repertoires, which include quadrupedal walking and running, leaping, and hopping (Gebo, 1987; Nash et al. 1989). However, the morphologies of these two species reflect different adaptations. In the galago, the calcaneus and navicular bones are extended toward the distal foot as a morphological adaptation for leaping (Hall-Craggs, 1965; Gebo, 1993). In contrast, the lemur lacks the specialised foot morphology of the galago and is considered a generalist with labile behaviour (Ward & Sussman, 1979; Crompton et al. 1987).
Demes et al. (1999) compared the kinetics of leaping between leapers and generalists according to the classification of Napier & Walker (1967), in which some prosimians were assigned to a vertical clinging and leaping group based on their positional behaviours and morphological characteristics. Although very few data are available for Garnett's galago, and most members of the leaper group are of the Propithecus genus, take-off force is higher in the generalist (including Lemur catta) than in a leaper of comparative body weight leaping the same distance between compliant poles designed to mimic natural conditions (Demes et al. 1999). Leaping distance depends on take-off force, acceleration distance, and take-off angle (Crompton et al. 1993; Demes et al. 1999). A leaper leaps a given distance effectively, with a long acceleration distance due to long hind limbs and with a low take-off force, whereas a generalist leaps the distance with a short acceleration distance and high take-off force (Demes et al. 1999). Although the locomotor repertoires of the lemur and galago do not differ greatly, given the suggestions by Demes et al. (1999) and other morphological studies of prosimian hind limbs (Hall-Craggs, 1965; Oxnard et al. 1981), the results of our study might reflect the difference between a leaper and a generalist, which means that the galago has developed a mechanical system enabling elongation of the acceleration distance or speed-oriented mechanics.
However, comparisons of species of different body weights, such as Galago garnettii and Lemur catta, must account for the scaling effect. The force during locomotion is proportional to body weight2/3; thus, the relative force (force/body weight) should be proportional to body weight−1/3 or follow negative allometry (Alexander, 1985; Demes et al. 1999). The smaller galago is expected to produce a higher force relative to the lemur, but the galago has speed-oriented ankle joint mechanics. This discrepancy might be mediated by muscle size and architecture. Unfortunately, no data on muscle weight or fiber arrangement in Galago garnettii and Lemur catta are available, but the more specialised leaper Galago senagalensis has been demonstrated to have relatively heavier superficial plantarflexors than those of quadrupedal primates (Demes et al. 1999) . Consequently, in terms of force production, the speed-oriented joint mechanics of the galago might be complemented by muscle size and/or other joint power. In the loris, the flexors of the digits account for most of the total lower leg muscle weight (Grand, 1967). This arrangement is consistent with the high MAs of digit flexors, which exacerbates the interspecific difference and confirms our identification of power-oriented ankle joint mechanics in the loris.