This article was published online on 6 June 2013. An error was subsequently identified. This notice is included in the online and print versions to indicate that both have been corrected 18 June 2013.
Correspondence to: Dr. Pierre Lemelin; Division of Anatomy, 5-05A Medical Sciences Building, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta T6G 2H7, Canada. Fax: (780) 492-0462. E-mail: email@example.com
Prosimian primates (strepsirrhines and tarsiers) are remarkably diverse in body size and locomotor behavior (Napier and Walker, 1967; Walker, 1974, 1979; Gebo, 1987; Oxnard et al., 1990; Smith and Jungers, 1997; Fleagle, 1999). The smallest prosimian (Microcebus myoxinus) weighs about 30 g, while the largest (Indri indri) can reach 5.8–6.8 kg (Smith and Jungers, 1997). Their wide-ranging locomotor behaviors are broadly classified into the categories of arboreal quadrupedalism, slow climbing, and vertical clinging and leaping (Napier and Walker, 1967; Walker, 1974, 1979; Gebo, 1987; Hamrick, 1996a-1996c; Fleagle, 1999). These varied behaviors have been correlated with hindlimb morphology, including quantitative aspects of the musculature. For example, leaping prosimians are characterized by greater mass of the quadriceps femoris muscle used for knee extension (41–49% of hindlimb muscle mass) compared to active arboreal quadrupeds like Varecia variegata (24% of hindlimb muscle mass) (Demes et al., 1998). Similarly, a much larger gastrocnemius muscle used for ankle plantarflexion is found in the leaping Galago (25% of extrinsic foot muscle mass) compared to slow climbing taxa such as Nycticebus and Perodicticus (5–12% of extrinsic foot muscle mass) (Grand, 1967; Jolly and Gorton, 1974; Gebo and Dagosto, 1988). The relative mass of the ankle plantaflexors also distinguishes leapers of different body size. The smaller, foot-powered jumpers (e.g., Galago senegalensis) have a relatively larger ankle plantarflexors compared to the larger, thigh-powered jumpers (e.g., Propithecus verreauxi) (Gebo and Dagosto, 1988; Demes et al., 1998). Moreover, Demes et al. (1998) found that the mass of the propulsive muscles of the hindlimb of prosimians scales with body mass at an exponent lower than required to maintain functional equivalence of force production (observed slope of 1.22 vs. expected slope of 1.5) between small- and large-bodied leapers.
While many facets of the prosimian hindlimb and its musculature have been quantified, the bulk of existing data on the musculature of the forelimb is restricted to detailed anatomical descriptions (e.g., Murie and Mivart, 1872; Miller 1932, 1943; Forster, 1934; Straus, 1941a, 1941b, 1942; Jouffroy, 1962, 1975; Schultz, 1984; Lewis, 1989). Despite the lack of quantitative data, several comparative observations on the relative size and development of the prosimian forearm and hand musculature have been made with functional inferences relating to grasping specializations, positional behavior, and body size. For example, Forster (1934) described the forearm digital flexor mass of Perodicticus as remarkably well developed, with the flexor digitorum profundus (FDP) muscle having two large and distinctive heads contributing to the pollex and digits III to V (with no contribution to the diminutive index finger). He considered the presence of these large digital flexor muscles, along with large flexor pollicis brevis and adductor pollicis muscles in the palm, to be a byproduct of the specialized, pincer-like arrangement of the hand in this slow climbing lorisid (i.e., wide divergence of the pollex accompanied with extreme reduction of the index finger; see Forster, 1934; Straus, 1942; Jouffroy and Lessertisseur, 1977, 1979; Lemelin and Jungers, 2007). In the same vein, Miller (1943) inferred the digital flexors to be much stronger than the extensors in slow climbing lorises and qualified both the FDP and ulnar deviators as large, whereas Napier and Walker (1967) and Walker (1974) describe the ulnar border of the hand of vertical clingers and slow climbers to be more dominant compared to that of arboreal quadrupeds. Although the carpal morphology of vertical clingers, and especially slow climbers, allows for greater ulnar deviation compared to other quadrupedal prosimians (Cartmill and Milton, 1977; Hamrick, 1996a-1996c; Lemelin and Schmitt, 1998), there is no quantitative evidence that carpal features enhancing ulnar deviation correlate with a relatively larger ulnar musculature. Ulnar deviation, which involves rotation and sliding of the hand at the wrist joints toward the ulnar side of the forearm (Wood-Jones, 1949), allows grasping of an arboreal substrate between the pollex and postaxial digits during the support phase of quadrupedal walking and climbing (Sarmiento, 1988; Preuschoft et al., 1993; Lemelin and Schmitt, 1998). With regard to scaling effects on the musculature of the forearm and hand, Napier and Walker (1967) and Walker (1974) also suggested that the pincer-like hands of large vertical clingers such as Propithecus and Indri require greater prehensile force to support their relatively larger body mass due to the increased forces of gravity and friction acting on the ulnar border of the hand while clinging.
Anatomical differences in relative development and size of the forearm and hand musculature of prosimians have been noted in past studies, and some of these differences have been linked to grasping specializations, positional behavior, and body size. These observations have never been substantiated with quantification of forearm and hand muscle weights, although recently, such data have been reported for the pectoral musculature of prosimian primates (Wright-Fitzgerald et al., 2010). Moreover, unlike the hindlimb musculature, little is known on how forelimb muscles of prosimians scale with body mass. To this end, we present the first quantitative study on the forearm and hand musculature of prosimians of different body size (58 g–3.6 kg) and locomotor adaptations (arboreal quadrupedalism, slow climbing, and vertical clinging and leaping). The aims of this study are: (1) to provide weights for individual muscles of the forearm and selected muscles of the hand in 12 prosimian species spanning over five strepsirrhine families and the genus Tarsius, (2) to document scaling relationships between forearm muscle mass and body mass among taxa of varying body size, and (3) to compare relative muscle mass by compartments and functional groups for the same sample. We also present qualitative differences in muscle anatomy between species, particularly with regard to the FDP muscle.
MATERIALS AND METHODS
The forearm and hand of 17 cadaveric specimens of prosimian primates were dissected (Table 1). All cadaveric specimens were obtained from the collections of the Duke Lemur Center (DLC), with the exception of one specimen of Otolemur garnettii belonging to Dr. Matt Ravosa (University of Notre Dame). All specimens were captive animals held in seminatural conditions that included free-ranging locomotion and posture. All specimens were adult, presented no obvious signs of degenerative diseases, and were all fresh-frozen. The right or left side was dissected depending on which side was best preserved. For all specimens (with the exception of O. garnettii), individual body masses prior to death were available from DLC records.
Table 1. Sample composition, locomotor categories, and individual body masses
We followed dissection and weighing methods prescribed by Jolly and Gorton (1974), Gebo (1985), Demes et al. (1998), and Wright-Fitzgerald et al. (2010) and adopted the muscle terminology used by Howell and Straus (1933) for the muscular anatomy of the rhesus macaque (Macaca mulatta). A typical dissection session started with thawing the specimen and lasted until each muscle was weighed and preserved in a 10% formalin solution, thus minimizing potential changes in muscle structure due to multiple freezing/thawing events. Using a small scalpel blade, forceps, and other dissection instruments, the skin and superficial fascia covering the forearm and hand were cut and completely removed from the limb. Particular attention was taken to remove the volar skin without damaging the underlying musculature and associated structures. Deep fascia covering muscle compartments was incised and removed. After gross description and photography of the specimen in different views, individual muscles from volar and dorsal compartments of the forearm and selected intrinsic muscles of the hand were carefully detached from their origin and insertion points and separated from all fascia and free tendon(s). Microdissection tools and a dissection microscope were used to dissect and prepare smaller muscles.
After being dissected and cleaned, fresh muscle samples were weighed three times to the nearest 0.01 g using a digital scale. The average of the three readings for each fresh muscle of each specimen is reported in Table 2. Then, each muscle was immersed for at least 48 h in a vial containing the formalin solution. Following this process, each muscle was removed from its vial, blotted dry on paper towel with minimal pressure, and weighed again three times. The average of the three weight readings for each formalin-fixed muscle of each specimen is reported in Table 3. We included this extra step, so that our muscle weight data can be more versatile for future studies as most specimens available from research collections are embalmed. All results from the scaling and relative weight analyses are reported for fresh-frozen muscle data only. Some intrinsic hand muscle weights were not reported for the smallest species because of their tiny mass (< 0.01 g). In some instances, those weights were grouped together as a single mass (thenar or hypothenar) as the muscle bellies were fused and not clearly distinguishable. Finally, some muscles were not observed or not sampled due to excessive dryness as a result of freezing (noted as absent or damaged in Tables 2 and 3).
Table 2. Individual mass (g) of forearm and selected hand muscles (fresh-frozen) for 17 prosimian specimens
Relationships between muscle mass and individual body mass obtained from DLC records were investigated using model II regression line-fitting methods as both mass variables exhibit biological variation, thus are measured with error, and are considered independent of each other (Rayner, 1985; Sokal and Rohlf, 2012). Reduced-major axis (RMA) slopes and associated 95% confidence intervals (CI) were computed and compared to theoretical slopes describing geometric similarity between muscle mass and body mass (i.e., a slope of 1.0 is indicative of shape preservation with changes in body mass). These scaling relationships were examined for total muscle mass of the forearm (volar and dorsal compartments) and for mass of the digital flexor muscles (i.e., flexor digitorum sublimis and flexor digitorum profundus) for all the 12 species. Average values were used when more than one specimen was available. Relative muscle mass as a percentage of total forearm muscle mass was quantified by compartment (volar and dorsal) and functional group for all the 12 species. Functional groups of forearm muscles included: (1) ulnar deviators of hand (extensor carpi ulnaris and flexor carpi ulnaris); (2) radial deviators of hand (extensor carpi radialis longus, extensor carpi radialis brevis, and flexor carpi radialis); (3) digital extensors (extensor digitorum communis and extensor digiti proprii); and (4) digital flexors (flexor digitorum sublimis and flexor digitorum profundus). Relative weights of selected intrinsic hand muscles were also quantified across a subset of seven larger prosimian taxa. Thenar and hypothenar muscles included abductor, flexor, and opponens (when present) muscles of the pollex and digit V. Relative muscle mass data were compared qualitatively between species as small sample sizes prevented the use of statistics for this portion of the study.
Scaling analyses of body mass with total forearm muscle mass (i.e., all muscles of the dorsal and volar compartments of forearm) or digital flexor muscle mass (i.e., flexor digitorum sublimis and flexor digitorum profundus) show very consistent results shown in Fig. 1. Total forearm and digital flexor muscle masses are highly correlated with body mass (r = 0.97 and r = 0.96) and scale with a slope of 1.15 and 1.20, respectively (Fig. 1). The lower bound of the 95% CI of those RMA slopes (0.95 and 0.98) includes the isometric slope value of 1.0. This suggests that as body mass increases, muscle mass of the forearm and digital flexors of prosimian primates increases geometrically. Of note is the outlier Tarsius well above the regression lines, particularly for the digital flexor muscle mass comparison (Fig. 1b). This suggests a relatively heavier forearm and digital flexor musculature in this small-bodied prosimian.
Quantitative data on volar (flexor) and dorsal (extensor) compartments of the forearm are shown in Fig. 2. The mass of the volar compartment represents between 52.2 and 65.8% of total forearm musculature in prosimians, while the mass of the dorsal compartment represents between and 34.2 and 47.8% of total forearm musculature. These ranges are unexpectedly narrow for such a diverse sample of primates. Furthermore, variation in compartment muscle mass does not sort according to predictions made on the basis of grasping specializations of the hand and locomotor adaptations. For example, the slow-climbing Nycticebus, with pincer-like hands, does not display greater development of the flexors relative to the extensors compared to other taxa (Fig. 2). The lower value of 56.3% for the volar compartment of Propithecus verreauxi is also surprising, especially when contrasted to the value of 65.8% found in Varecia, an arboreal quadruped of similar body size. However, the well-developed brachioradialis muscle (part of the dorsal compartment and potential flexor of the forearm at the elbow joint) represents 16.3% of total forearm muscle mass in Propithecus, which is more than twice the average of 6.96% for other prosimian species. If the brachioradialis muscle is removed from the overall forearm mass calculation, the relative mass of the volar compartment of this large-bodied, vertical clinger jumps to 70.6%. As a general trend within the prosimian sample, the relative mass of the volar compartment gradually increases as body mass goes up: 52.2% in Microcebus to 70.6% in Propithecus (excluding the brachioradialis muscle). A similar trend with body mass increase can be noted among more closely related cheirogaleids (52.2% in Microcebus, 55.8% in Cheirogaleus, and 60.5% in Mirza).
Quantitative data of relative muscle mass by functional group (i.e., ulnar deviators of hand, radial deviators of hand, digital flexors, and digital extensors) are shown in Fig. 3. Relative muscle masses by functional grouping do not vary in any predictable manner among locomotor categories. For example, the ulnar deviators of vertical clingers and slow climbers are not more developed compared to those of active arboreal quadrupeds of similar size. In fact, the largest ulnar deviators are found in Microcebus (20.8% of total forearm mass) and the smallest are in Propithecus (11.6% of total forearm mass, Fig. 3). Similarly, the digital flexors of Propithecus and Nycticebus are not more developed compared to those of arboreal quadrupeds of similar size (i.e., Varecia and Otolemur).
Quantitative data for selected intrinsic hand muscles for a subsample of larger prosimians are presented in Fig. 4. The relative mass of the adductor pollicis muscle in Nycticebus (51.6% of total intrinsic muscle mass) is markedly higher compared to all other taxa. The large adductor pollicis muscle mass of Nycticebus confirms qualitative observations made by Forster (1934) in Perodicticus and underscores the potential functional importance of this muscle in the gripping mechanism of the pincer-like hand of lorises.
This study presents the first quantitative data on the musculature of the forearm and hand of prosimian primates. Our analyses suggest isometric scaling of the forearm and digital flexor musculature relative to body mass, with respective RMA slope values of 1.15 and 1.20. In other words, muscle mass of the forearm and digital flexors of prosimian primates increases geometrically with body mass. Slopes obtained in this study are similar to those reported by Demes et al. (1998) for hindlimb muscle mass of prosimian primates. Other scaling studies of primate limb musculature have reported similar scaling relationships. For example, flexor muscles of the forearm in six primates of widely different size and phylogenetic affinities (Otolemur, Cercopithecus, Colobus, Papio, and Homo) scale isometrically (with a slope of 1.19 ± 0.21 95% CI) when scaled to body mass (Alexander et al., 1981). However, in a comparison involving a narrower size range of closely related taxa, Myatt et al. (2012) reported negative allometry of the forearm digital flexor muscle mass with body mass (with a slope of 0.71) for a sample of orangutans and African apes. These examples are reminders that slopes values are only descriptors of size-related phenomena that require further explanations and can vary depending on the phylogenetic structure and composition of the sample being compared (Giles, 1956; Smith, 1980; Jungers, 1984, 1985; Jungers et al., 1995).
Yet, the slope values for muscle mass relative to body mass presented earlier, including those for the forearm muscle mass of prosimian primates, all fall short of the predicted slope value of 1.5 if muscle force (proportional to cross-sectional area) increases isometrically with body mass (Demes et al., 1998). In other words, muscle force decreases steadily as body mass increases. All other factors being equal and in the absence of information on fiber length in these muscles, it is tempting to conclude that the hands of larger prosmian primates exert less prehensile forces per unit of body mass compared to smaller ones. But not all factors are always equal. For example, the digital flexors of great apes are characterized by having the longest muscle fibers and highest physiological cross-sectional area of all muscles of the forearm (Myatt et al., 2012). This kind of architecture is likely to provide functional advantages for muscles requiring both force and excursion as the digits and hand assume many different positions during prehensive behavior (Alexander et al., 1981; Myatt et al., 2012). In large-bodied prosimians such as Propithecus and Indri, the pollical and ulnar metacarpals display positive allometry, which contribute to a wider span of the digits when grasping a branch (Lemelin and Jungers, 2007). A wider span of the hand allows the normal component of the adductor force of opposing digits to increase (as the subtended angle approaches 180°) when grasping a branch (Cartmill, 1985). These examples illustrate how larger primates may be able to compensate for this deficit in relative muscle mass (and grasping force) by evolving other aspects of limb anatomy.
The departure of Tarsius from the regression lines in the scaling analyses of muscle mass requires further explanation. Although caution should be exercised with examination of residuals in regression analysis (Jungers et al., 1995), Tarsius is positioned well above the regression line when compared to prosimian taxa of similar body mass (i.e., narrow allometry comparison). The relatively larger forearm muscle mass of Tarsius, particularly the digital flexor musculature, covaries with very elongated fingers relative to body mass (Lemelin and Jungers, 2007). The disproportionally long fingers of Tarsius are used to catch fast-moving insects (Niemitz, 1984; Lemelin, 1996), which in turn may require bulkier flexor muscles when grasping and securing prey.
Analyses of relative muscle mass by forearm compartment and functional group yielded mixed results. As body size increases among prosimians, so does the mass of the volar (flexor) muscle compartment relative to the dorsal (extensor) compartment. A narrow allometric comparison involving cheirogaleids (58–240 g) reveals similar body size-related changes. Comparisons of the relative mass of the forearm and hand muscles by functional group did not identify any trend associated with locomotor adaptations or grasping specializations of the hand. The ulnar deviators of the hand are not more developed in taxa emphasizing ulnar deviation of the hand such as vertical clingers and slow climbers compared to arboreal quadrupeds with less deviated hand postures (Hamrick, 1996a, 1996b; Lemelin and Schmitt, 1998). For example, the relative mass of flexor carpi ulnaris and extensor carpi ulnaris muscles is smaller in vertical clingers when compared to arboreal quadrupeds of similar body size (Tarsius and Galago vs. Microcebus and Cheirogaleus; Propithecus vs. Varecia) or close phylogenetic affinities (Hapalemur vs. Eulemur). The ulnar deviators of the slow climbing Nycticebus are not notably larger than those of the arboreal quadruped Otolemur garnettii, another member of the superfamily Lorisoidea of similar size as the slow loris. Likewise, the digital flexors (flexor digitorum sublimis and flexor digitorum profundus) are not more developed in prosimians with pincer-like hands (Nycticebus and Propithecus) compared to other taxa. For example, the digital flexors vary little between two vertical clingers of very different body size (49.6% of forearm muscle mass in the small-bodied Tarsius and 52% in the large-bodied Propithecus). At 45.9%, the relative mass of the digital flexors of Nycticebus falls right within the middle of the range of our prosimian sample (35.4% in Microcebus to 52% in Propithecus).
The lack of clear functional associations between relative muscle mass and hand use during positional behavior is surprising for a group of primates as diverse as prosimians. In contrast, studies of the prosimian hindlimb have established clear correlates between increase in relative mass of some muscle groups (e.g., quadriceps femoris or gastrocnemius) and emphasis on leaping behavior (Grand, 1967; Jolly and Gorton, 1974; Gebo and Dagosto, 1988; Demes et al., 1998). This does not seem to be the case for the forearm and hand, with the exception of the relatively more developed adductor pollicis muscle in Nycticebus. No experimental evidence exists that can elucidate the functional role of this pollical muscle during grasping; however, such evidence is well-documented for the adductor hallucis muscle, its pedal homolog. The adductor hallucis of Nycticebus shows its highest levels of activity when the foot grasps small dowels during activities lacking weight bearing such as reaching or bridging (Kingston et al., 2010). As the hands and feet of Nycticebus and Perodicticus are remarkably similar in overall shape and proportions (Forster, 1934; Jouffroy and Lessertisseur, 1979; Lemelin and Jungers, 2007), hence the term “quadrumanous” often associated with the locomotion to these primates, it is likely that the functional role of adductor pollicis may be similar to that of adductor hallucis.
The well-developed adductor pollicis muscle is not the only feature that distinguishes Nycticebus from other prosimian primates. Although the mass of flexor digitorum profundus (FDP) muscle of Nycticebus is unremarkable, the architecture and arrangement of the long digital tendons are very different. As shown in Fig. 5, the FDP muscle of Nycticebus is different compared to other prosimians in having two very distinctive heads and more individualized tendons, a condition also reported in other lorises (Murie and Mivart, 1872; Forster, 1934; Straus, 1942). One head on the radial side arises from the shaft of the radius (corresponding to the radialis and condyloradialis portions of the FDP) and the other on the ulnar side arises from the medial epicondyle of the humerus and shaft of the ulna (corresponding to the condyloulnaris portion of the FDP). The radial head sends individual tendons to digits I–IV and ulnar head to digits I, IV, and V (Fig. 5). In other prosimians such as Propithecus shown in Fig. 5, the FDP has usually three interconnected heads (radialis, condyloradialis, and condyloulnaris) sending tendons that will fuse to varrying degree into a common lamina at the level of the carpal tunnel (Murie and Mivart, 1872; Straus, 1942; Jouffroy, 1962). From this lamina, individual tendons will be given to each digit (Fig. 5). The functional consequences of such differences in muscle and tendon morphology remain unexplained. Experimental data from other vertebrates suggest that stretching qualities of more individualized tendons would expand the functional range of muscles, allowing muscle fibers to remain within their optimal force-length ratio as joint angular excursion varies with different locomotor activities (Roberts, 2002). In this regard, flexion of the digits in Nycticebus and other lorises occur within a wide range of hand postures, including flexion, extension, and ulnar deviation at the wrist joints (Dykyj, 1980; Hamrick, 1996a; Lemelin and Schmitt, 1998). It is also important to point out that the pollex and digit IV of Nycticebus, the longest components of the pincer mechanism opposing each other during grasping, are the only rays receiving contributions from both muscle heads. Although no electromyographical evidence exists for the FDP of Nycticebus, Kingston et al. (2010) reported the highest levels of activity in one of the crural homologs of FDP (flexor digitorum tibialis) during locomotor behaviors requiring power grasping of the foot. Relying on the same analogy as above on the quadrumanous nature of the locomotor adaptations of Nycticebus, it is possible that the functional role of FDP may be the same.
Quantitative data were presented for the musculature of the forearm and hand of 12 prosimian species. Isometric scaling of muscle mass relative to body mass was found for the forearm musculature and digital flexor muscles. Furthermore, interspecific comparisons of muscle weights by compartment and functional groups could not identify any association between relative development of muscles that move the hand and digits and locomotor adaptations or grasping specializations of the hand. One exception is the hand of Nycticebus with its widely divergent pollex correlating with a well-developed adductor pollicis muscle and a peculiar muscle-tendon arrangement of the flexor digitorum profundus muscle. Future research should focus on muscle fiber architecture, electromyography, and sonometry of these muscles in order to better understand the functional morphology and evolution of the pincer-like hand of lorises.
The authors are very grateful to the Duke Lemur Center (Durham, NC) for giving permission to use cadaveric material under their care, Dr. Ken Glander, David Haring, and David Brewer for facilitating access to this invaluable collection, and Dr. Sarah Zehr for providing the body mass data for the specimens. The authors are also indebted to Drs. Brigitte Demes, Matt Ravosa, and Pat Wright for giving access to galago, lemur, and tarsier specimens, the Department of Evolutionary Anthropology (Duke University) and Department of Anatomical Sciences (Stony Brook University) for providing laboratory space and access to equipment, and to Drs. Daniel Schmitt and Christine Wall for hospitality and friendship while in North Carolina. Dr. Christine Wall and two anonymous reviewers provided very useful comments on an earlier draft of the manuscript. This is DLC publication #1246.