Architectural and histochemical diversity within the quadriceps femoris of the brown lemur (Lemur fulvus).код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 69:355-375 (1986) Architectural and Histochemical Diversity Within the Quadriceps Femoris of the Brown Lemur (Lemur fulvus) FREDc. ANAPOL AND WILLIAM L. JUNGERS Departments of Oral Anatomy and Orthodontics, College ofDentistry, University of Illinois at Chicago, Chicago, Illinois 60612 (I? C.A.); Department of Anatomical Sciences, School of Medicine, State University of New York at Stony Brook, Long Island, New York 11794 WL.J.) KEY WORDS Fiber architecture, Muscle histochemistry, Lemur fuluus, Quadriceps femoris ABSTRACT Physiologically related features of muscle morphology are considered with regard to functional adaptation for locomotor and postural behavior in the brown lemur (Lemur fuluus). Reduced physiological cross-sectional area, estimated maximum excursion of the tendon of insertion, length of tendon per muscle fasciculus, and areal fiber type composition were examined in the quadriceps femoris in order to assess the extent of a “division of labor” among four apparent synergists. Each of these four muscles in this prosimian primate displays a distinguishing constellation of morphological features that implies functional specialization during posture and normal locomotion (walWrun, galloping, leaping). Vastus medialis is best suited for rapid whole muscle recruitment and may be reserved for relatively vigorous activities such as galloping and leaping (e.g., small cross-sectional area per mass, long excursion, predominance of fast-low oxidative fibers, relatively little tendon per fasciculus). In theory, rectus femoris could be employed isometrically in order to store elastic strain energy during all phasic activities (e.g., large cross-sectional area per mass, short excursion, predominance of fast-high oxidative fibers, large amount of tendon per fasciculus). Vastus intermedius exhibits an overall morphology indicative of a typical postural muscle (e.g., substantial cross-sectional area, short excursion, predominance of slow-high oxidative fibers, large amount of tendon per fasciculus). The construction of vastus lateralis reflects an adaptation for high force, relatively high velocity, and resistance to fatigue (e.g., large crosssectional area, long excursion, most heterogeneous distribution of fiber types, large amount of tendon per fasciculus); this muscle is probably the primary contributor to a wide range of locomotor behaviors in lemurs. Marked dramatic architectural disparity among the four bellies, coupled with relative overall fiber type heterogeneity, suggests the potential for exceptional flexibility in muscle recruitment within this mass. One interpretation of this relatively complex neuromuscular organization in the brown lemur is that it represents an adaptation for the exploitation of a three-dimensional arboreal environment by rapid quadrupedalism and leaping among irregular and spatially disordered substrates. The analysis of muscle function in primate locomotion is complicated by the vast architectural and histochemical variability that is found among and within whole muscles. Differences in gross architectural Configuration 0 1986 ALAN R. LISS, INC. can affect contractile properties of muscles such as tetanic tension and maximum velocity (Gans and Bock, 1965; Gans, 1982).Other Received April 19,1985; revision accepted September 13,1985. 356 F.C. ANAPOL AND W.L. JUNGERS properties, such as time-to-peak-tension and “fatigability,” appear to be the consequence of cytological or histochemical differences among muscle cells (Denny-Brown, 1929; Burke et al., 1971; Burke, 1981). Much of the prior work on the relationship between muscle form and function has emphasized specific aspects of muscle morphology and how myological variation is associated with differences in locomotor behavior observed among closely related taxa. These studies include, for example, relative distribution of muscle mass (Ashton and Oxnard, 1963; Stern, 1971; Fleagle, 19771, the proportion of tendon composing whole muscles (Alexander, 1974; Alexander and Bennet-Clark, 1977; Biewener et al., 19811, and relative fiber type composition (Sickles and Pinkstaff, 1981). Recent in vivo studies further demonstrate how architectural and histochemical differences among muscles can account for differences in the contractile properties of whole muscles as well as how intra- and intermuscular recruitment occurs during mammalian locomotion (e.g., Gillespie et al., 1974; Armstrong et al., 1977; Goslow et al., 1977; Smith et al., 1977; Sullivan and Armstrong, 1978; Taylor, 1978; Walmsley et al., 1978; Smith et al., 1980; Spector et al., 1980; Walmsley and Proske, 1981; Anapol and Jungers, 1982; Gardiner et al., 1982; Muhl, 1982). The manner in which the wide variety of myological configurations that can be generated by these several variables contributes to functional specialization among muscles remains unclear. For any taxon, the allocation of the “labors of locomotion” among muscles of a synergistic group may reflect relatively subtle, yet genuinely distinctive, requirements for positional behavior. Differences between phylogenetically disparate species that appear to exhibit similar locomotor repertoires are especially critical in the recognition of less-obvious requisite differences in adaptation to the environmental substrate. In this study of quadriceps femoris in Lemur fuluus, we demonstrate that vastus medialis (VM), rectus femoris UXF), vastus intermedius (VI), and vastus lateralis (VL) are each characterized by a unique constellation of physiologically related morphological features, implying fuctional specialization for each muscle of this synergistic group. Architectural disparity among the four muscles, coupled with overall fiber type heterogeneity, suggests functional versatility for optimal exploitation of a complex arboreal environment via a locomotor repertoire dominated by rapid quadrupedalism and leaping among substrates of various sizes and spatial arrangements Petter et al., 1977; Ward and Sussman, 1979). MATERIALS AND METHODS Morphology Architectural morphology was studied in five adult brown lemurs (Lemur fuluus). Two specimens were obtained from the collection of the Department of Anatomical Sciences, State University of New York at Stony Brook. Three additional specimens were available after the muscles of the right hindlimb were removed for histochemical analysis of fiber type distribution (see below). All were thawed and fixed by submersion ( 7 4 % formaldehyde, 3% phenol) with one hindlimb set in a typical quadrupedal postural position as estimated from videotapes and personal observation of live animals (Fig. 1). A total of four VM and RF and five VI and VL were examined. Each muscle was identified, following Murie and Mivart (1872), Jo&roy (1962),and Jungers et al. (19801, and dissected free from its attachments to the bone. METATARSOTIBIO-TALAR METATARSAL Fig. 1. Approximate postural angles of hindlimbjoints in Lemur fuluus. H, horizontal;arrow points anteriorly. 357 ARCHITECTURE AND HISTOCHEMISTRYIN LEMUR QUADRICEPS FEMORIS sured between the proximal myotendinous junction (MTJ)and the distal MTJ; acute angle of pinnation (Q) with respect to the probable line of action of the muscle (see above); length of tendon extending from the proximal MTJ of the fasciculus to its proximal attachment to bone ( 1; and length of tendon extending from the istal MTJ of the same fasciculus to its distal attachment to the bone 2 (td). Fasciculus, proximal tendon, and distal tendon measurements for the parallel-fibered VM were taken along the estimated line of action and along the medial and lateral margins of the muscle and were averaged to generate a single value. The muscles were trimmed of free tendon, blotted dry,and weighed on a Mettler (H54AR) balance scale. Reduced physiological cross-sectional area was calculated for RF,VI, and VL by using the formula (Schumacher, 1961; after Weber, 1851; and adjusted by Haxton, 1944) Fig. 2. Measurement protocol for pinnate muscles; If, reduced physiological cross-sectional area(cm2) = [mass (gm) x cos Q]! (1) [If (cm) x specific density] tp, td, and 8 are explained in Material and Methods; x and y are normal to lf and represent portions of the true physiologicalcross-sectionalarea of the whole muscle. Gross measurements of the maximum length of the muscle belly and the maximum breadth of the belly perpendicular to the length facilitated choice of sampling sites. For architectural measurements in pinnate muscles, bellies were sampled at the normal intersections of longitudinal' and transverse planes approximately 1 cm apart. Accordingly, the number of sampling points depended upon the length and width of the muscle belly. The longitudinal planes more or less bisected the smaller muscles and trisected the larger muscles. The transverse planes either trisected or quadrisected all bellies. At each sample point, six neighboring fasciculi were measured--three proximal and three distal to the point; for each variable, the mean of the six was used in subsequent calculations. The following parameters were recorded (Fig. 2): length of fasciculus (lf)mea'Longitudinal planes were parallel to the probable line of action of the whole muscle defined as a straight line passing between the midpoints of the proximal and distal attachments (followingStern, 1971). where the specificdensity of muscle is 1.0564 gm/cm3 (Murphy and Beardsley, 1974). Because the fibers insert into the tendon at an angle to the direction that the whole muscle contracts, the "useful vector" is resolved by including cos Q in the calculation (Haxton, 1944; Gans, 1982). This is ignored in the calculation for parallel-fibered muscles where COSQ= 1. Estimated maximum excursion of the tendon of insertion (h) was calculated from lf and 0 according to the following equation (adapted from Benninghoff and Rollhauser, 1952) h = If (cos 0 - dcos" 0 + n' - 1). (2) The maximum coefficient of contraction (n) (where n = the length of a fiber after contractiodthe length of a fiber a t rest) was assumed to be 70%(Gans and Bock, 1965). For each muscle, the ratio of the mean total tendon length (tt = tp + td) to the mean of fasciculus length plus total tendon length was calculated. The proportion of mass to predicted effective maximal tetanic tension (Po)was calculated to compare the priority of muscle force versus muscle velocity for a given mass (fol- 358 F.C. ANAPOL AND W.L. JLJNGERS lowing Sacks and Roy, 19821, where Po = reduced physiological cross-sectional area (cm') x specific tension of muscle [assumed to be approximately 2.3 kg/cm2 (Spector et al., 1980)l. Measurements and calculations for each muscle were averaged for four (VM, RF) or five (VI, VL)animals and standard errors of the means were computed. For each variable, coefficients of variation adjusted for small sample size (V") (Sokal and Rohlf, 1981)were calculated among the means of the four muscles and are expressed k standard error k,*). Histochemistry Three adult male brown lemurs (L. fuluus subspecific hybrids: 7 years12.25 kg, 3 years1 1.9 kg, 3 years12.15 kg) were killed by cardiac injection of sodium pentobarbital 20 minutes after sedation by intramuscular injection of Ketaset. Muscle tissue was removed from the centermost portion of each belly and quenched in isopentane (2-methyl-butane) previously cooled to approximately - 150°C with liquid nitrogen. After 20-25 seconds, the tissue was quickly removed to a straight liquid nitrogen bath. Histochemical treatments were performed concurrently and closely followed standard protocols for alkaline and acid myosin adenosine triphosphatase (ATPase) (Guth and Samaha, 1970), nicotinamide adenosine dehydrogenase-tetrazolium reductase (NADH-TR), and succinate dehydrogenase (SDH) (Dubowitz and Brooke, 1973). Because the treatment for myosin ATPase is pH dependent and variation in pH sensitivity is found among species (Brooke and Kaiser, 1970), optimal preincubation environments were determined for lemur muscle. Following preincubation a t pH 10.35, three levels of staining intensity or myosin ATPase activity were clearly distinguishable: light (low activity), dark (high activity), and a n intermediate level that appeared slightly lighter than the dark-staining fibers. The low-activity fibers were presumed to be slow contracting fibers and the high- and intermediate-activity fibers were presumed to be fast contracting (Guth and Samaha, 1970; Burke et al., 1971; Peter et al., 1972).Following preincubation at pH 4.35, slow fibers exhibited high activity while the activity of the fast fibers appeared to be inhibited. When sections are treated for NADH-TR or SDH, three levels of staining intensity again are distinguishable (following Stein and Padykula, 1962; Henneman and Olson, 1965): light (low concentration of reactive product), medium, and dark. However, these do not necessarily correspond to the three cell types that are present following alkaline myosin ATPase treatment (Nemeth and Pette, 1981). For the current study, cells that exhibited medium and high concentrations of oxidative enzyme were designated highly oxidative (0). The quantification of fiber type composition proceeded in the following manner. Two consecutive sections (one treated for alkaline-stable myosin ATPase and the other for NADH-TR)were examined under a light microscope to determine whether corresponding groups of cells were suitable for analysis. Sections treated for acid-stable myosin ATPase and SDH could be substituted, if necessary, or used to substantiate identifications. For each alkaline-preincubated myosin ATPase section, deep, middle, and superficial regions at varying mediolateral locations were selected. A field was projected onto a Sharp Linytron-plus color monitor receiver (XR-3019)through a Sony Vititron black-andwhite video camera mounted on a Zeiss (West Germany) Universal microscope that was fitted with a planapo objective. The monitor was interfaced with a Tektronix (4632) video hard copy device. A hard copy was made of the region of choice and the cells were identified (either by observation of the monitor or through the eyepiece of the microscope) and marked slow or fast according to the intensity of myosin ATPase activity. The corresponding region on the NADH-TR slide was located and the same cells were marked 0 if highly oxidative. Superimposing these two classificatory systems, three classes of cells were tabulated slow twitch-high oxidative (S), fast twitchhigh oxidative (FO), and fast twitch-low oxidative (F). Cross-sectional areas were measured for six cells (with intact boundaries) from each class represented within a field with a GrafPen sonic digitizer (Scientific Accessories Corporation). In nonoblique sections, areas of cells were measured directly in square centimeters. In oblique sections, the smallest diameter of a cell was measured as a precaution against distortion (Dubowitz and Brooke, 1973) and converted to area [A = pi x (d/2?]. Direct and indirect methods of area determination were never mixed 359 ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS within a single field. The mean areal values for each class were converted to square micrometers, according to scale. For each field, the number of cells in each class was multiplied by the average crosssectional area per fiber for that class. These lcm products were summed and each product was divided by the sum of products to yield the percentage of cross-sectional area for a class. For each class of fibers, the percentages of cross-sectional area for all three fields from each section were averaged. If more than one sample was taken from a muscle, the samples were averaged. For each class, the three animals were averaged to determine the value for a “composite” animal. For each class, two-factorial analysis of variance (Sokal and Rohlf, 1981)was applied to test the overall significance of differences Fig. 4. Rectus femoris (see Fig. 3 legend). Middle: among animals and among muscles. The coronal plane. Mann-WhitneyU-test (Sokal and Rohlf, 1981) tested for significance of differences between all possible pairs of muscles for each class of fibers in the composite animal. For each class the belly one-third to one-half of its length of fibers, V* f sv*was also computed for the distally. The more anterior portion of the attachment is fleshy. Distally, the belly tafour muscles. pers slightly and is truncated before continRESULTS uing as a flat tendon into the true patella. Gross morphology Some of the more anterior muscle fibers atVastus medialis (VM)(Fig. 3) arises from a tach directly to the superior patella (Jungers circumscribed attachment on the anterior as- et al., 1980) although a small percentage of pect of the femoral shaft, just distal to the the most anterior of these coalesce with the head. The most posterior portion of this at- major tendon of insertion. VM is a parallel-fibered muscle but at first tachment is by tendinous fibers which join glance it appears to be pinnate due to the irregular geometry of the belly. The long fasciculi are oriented from posteroproximal to anterodistal, which presumably corresponds to the direction of contraction of the whole muscle. VM most closely resembles the variety of parallel-fibered muscle described in Figure 2d in Gans (1982, p. 169). Although the long sides of VM are more or less parallel, the ends of the belly are not. This is because the fasciculi are not all of Ilcm equal length and serial replacement of muscle fiber by less-bulky tendon gradually reduces the anteroposterior dimension of the belly at both proximal and distal ends. The proximal tendon slightly covers the posterior margin of the proximal four-fifthsof the belly, and the distal tendon covers the deep surface of the anterior margin of the distal two-thirds of the belly (Fig. 3). Fig. 3. Vastus medialis, left hindlimb (L. fuluus). Top: Rectus femoris (RF) (Fig. 4) is a long, tasuperficial surface; bottom: deep surface; proximal attachment is to the right. pered cylindrical muscle incompletely 1 360 F.C. ANAPOL AND W.L. JUNGERS sheathed with thick tendon around the distal two-thirds of the belly. Proximally, it arises as two separate tendinous elements (not pictured) from the anterior inferior iliac spine and the cranial margin of the acetabulum (Jouffroy, 1962). These immediately merge into a single, round tendon which enters the proximal end of the belly. This proximal tendon becomes a flattened central tendon throughout most of the length of the muscle belly. The fasciculi radiate distally at an angle from the central tendon and continue into a tendinous sheath which envelops all but the most proximal surface of the muscle belly. As the belly tapers toward its distal termination, the tendinous sheath continues as a thick, narrow, flat tendon that passes superficial to the superior patella and vastus intermedius to insert into the true patella. In coronal section (Fig. 4),RF is bipinnate for most of its length. However, the fibers that originate from the distal end of the central tendon continue in more-or-lessthe same axis (as this tendon) as they proceed toward the tapered end of the belly and into the enveloping outer tendinous sheath. The fleshy fibers of vastus intermedius (vI) (Fig. 5 ) arise directly from the anterior periosteum of the femoral shaft with no intervening tendon. In sagittal section, the relatively short, unipinnate fasciculi angle distally from the femur and terminate in a dense superficial tendon that covers all but the most proximal end of the belly. The extrinsic portion of this tendon contains both the superior and true patellae and crosses the knee joint to attach into the tibia1 tuberosity. Vastus lateralis (VL) (Fig. 6 ) is a thick, fusiform muscle that arises from the anterior a Fig. 6. Vastus lateralis (see Fig. 3 legend). Middle: sagittal plane. crest of the greater trochanter. This attachment is composed of both the ends ofthe most proximal muscle fibers and the end of a thick superficial tendon that covers three-fourths of the proximal surface of the belly. In sagittal section, unipinnately arranged fasciculi arise from the superficial tendon and angle distally toward a thick deep tendon that underlies all but the most proximal end of the belly. TABLE 1. Sample sizes from lemur for architectural data (fasciculus length, angle ofpinnatwn, tendon length)' Specimen 1 2 3 4 5 Total Fig. 5. Vastus intermedius (see Fig. 3 legend). Bottom: sagittal plane. VM 3 3 3 3 12 RF VI VL Total - 36 18 18 18 18 108 54 54 54 54 54 270 90 123 105 105 99 522 48 30 30 24 132 'VM,vastus medialis; RF, rectus femoris; VI, vastus intermedius; VL, vastus lateralis. ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS 361 Distally, the belly ends abruptly as the deep tendon continues to its attachment into the true patella in common with the remaining quadriceps. Some fibers have been observed to insert into the superior patella (Jungers et al., 1980). Muscle architecture Sample sizes for whole muscles and fasciculi within muscles that were examined are presented in Table 1.Means (k standard error of the mean) for the raw measurements and calculations are also presented for each muscle in Table 2. Mean fasciculus lengths range from 1.77 cm in RF to 8.29 cm in VM. The angles of pinnation are similar among the three pinnate muscles (RF, VI, VL). More than a fivefold difference in mass separates the lightest muscle (VI) from the heaviest (VL). The largest average tendon length per fasciculus is found in the bipinnate RF while the smallest is in the parallel-fibered VM. The reduced physiological cross-sectional area accruing to each muscle is also expressed as a percentage of the total quadriceps area. VL is clearly the dominant muscle, contributing 58.4% of the estimated maximum potential force of the quadriceps. VM contributes the least to potential force, only 5.8%of the total. Despite its low mass, pinnation allows the potential force contribution of VI to attain 2.5 times that of the parallelfibered VM. The bipinnate muscle (RF) obtains only a one-third-larger value than does VI. Thus, pinnation does not necessarily increase potential force in all cases. With regard to the calculation of maximum excursion, the wide disparity between VM and the rest of the quadriceps suggests its relative specialization for high velocity (Table 2). Although much less dramatic than the divergence of VM, VL obtains 1.5 times the maximum excursion of RF and 1.4 times that of VI. The fiber architecture of VL may indicate a compromise of force and velocity within a single muscle (see Discussion). The greater mass of VL is due not only to a larger number of fibers, but also to the presence of longer fasciculi which would increase its maximum velocity (i.e., excursion/ time). This is indicated by the ratio of wet weight to predicted maximal tetanic tension @%/PO), which was calculated following Sacks and Roy (1982) (Table 2). They used this parameter as a means of estimating the priority of force versus length-velocity for each 362 F.C. ANAPOL AND W.L. JUNGERS hindlimb muscle in cat and interpreted high ratios as indicating maximization of lengthvelocity. In the lemur, the highest (by nearly threefold) value is seen in the parallel-fibered VM. Of interest is that the pinnatefibered muscle which monopolizes the force output of the group (VL) also obtains a relatively high length-velocity priority. The values for RF and VI are equally low and indicate priority of muscle force over lengthvelocity (for further discussion, see Sacks and Roy, 1982). The ratio of mean tendon length (tJ to the mean fasciculus length plus G (i.e.,&t tJ is also presented in Table 2. Again, the parallel-fibered VM is distinct from the three pinnate muscles. The two unipinnate mus- + cles (VI and VL) have similar values but both are less than that of the bipinnate RF. Despite the fact that the fasciculi of VL are continuous with tendon at both ends, when compared to VI (in which the proximal end of each fasciculus arises from bone), the longer fasciculi in VL neutralize any expected increase over VI in this calculation. Of the quadriceps, RF clearly has the “longest” tendon per muscle fasciculus. Consequently, each fasciculus should expend less energy (ATP hydrolysis) to transmit force to bone than if its contractile tissue were to extend the entire length of the muscle. VL and VI have the second highest proportion and although relatively close to RF’,the two one-joint muscles are clearly different from Fig. 7. Vastus medialis. A. ATPase preincubation at pH 10.35; B. ATPase preincubation at pH 4.35; C. NADH-TR D. SDH. S, slow twitch-high oxidative fiber; FQ, fast twitch-high oxidative fibers; F, fast twitch-low oxidative fiber. ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS RF. At the other end of the spectrum, VM has the least amount of tendon per fasciculus. Histochemistry Light micrographs from cross sections of VM, RF, VI, and VL are presented in Figures 7-10. In each group, four serial sections that have been treated for myosin ATPase (preincubation at pH 10.35 and 4.4),NADH-TR, and SDH have the corresponding cells in each treatment labeled as to class. In VM,RF, and VL, two cells have been identified as FOfrom each of the intermediate and darkstaining variety following treatment for alkaline myosin ATPase. These are thought to correspond to the fast oxidative glycolytic (FOG) and fast intermediate P i n t ) fibers in the tetrapartite classification of fiber types (Burke, 1975;McDonagh et al., 1980). 363 All three major cell classes, as well as all three levels of alkaline myosin ATPase reactivity, are represented in VM, RF, and VL. In VI, no F fibers are present and only slowand intermediate-staining fast fibers are observed. Also in VI, only fibers which stained the darkest and intermediate levels for oxidative enzyme are present, as observed in cells labeled S and FO in treatments for both NADH-TR and SDH. The means and ranges of fiber cross-sectional areas for each class of fibers are presented in Table 3 for individual animals. In every muscle, fiber size is generally correlated to increasing reactivity of myosin ATPase and decreasing presence of oxidative enzyme. Of interest is the extent to which the ranges of some (but not all) classes overlap. Fig. 8. Rectus femoris (see Fig. 7 legend). 364 F.C. ANAPOL AND W.L. JUNGERS Fig. 9. Vastus intermedius (see Fig. 7 legend). In Table 3, the percentage of cross-sectional area of each functional class of fiber type (S, FO, F) is included for three lemurs. The means and ranges of the three animals combined (composite lemur) are also indicated for each class and depicted graphically in Figure 11. A two-factorial analysis-of-variance table is presented in Table 4. For S and F classes of fibers, significant variability (P < .05) exists among muscles but not among animals. Neither the overall variability among muscles nor among animals is significant for the FO class. In other words, the animals are similar but the fiber-type properties of each muscle are variable. The Mann-Whitney U-test was employed a s a nonparametric test of significant differ- ences between all possible pairs of muscles for each class of fibers (Table 5). For S fibers, only VM and RF were not significantly different from one another (P < .05). For FO fibers, VM and VI were not significantly different from one another, nor were RF and VL. For F fibers, the only pairs not significantly different in their percentage of crosssectional area were RF and VL. In all three animals, VI is the slowest and most highly oxidative of the four muscles. The most extreme example of this is observed in lemur 2, which is composed of 100% S fibers. Of interest is the significantly higher percentage of cross-sectional area occupied by type S fibers in VL, when compared to VM or RF. Although all three are predominated by fast fibers (FO plus F) VM has a ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS 365 Fig. 10. Vastus lateralis (see Fig. 7 legend). significantly smaller percentage of FO fibers with a higher concentration of fast, low-oxithan either RF or VL. VM is the fastest and dative fibers (F)toward midbelly. the least oxidative of the four muscles, obDISCUSSION taining the highest percentage of F fibers The research design embodied in this work and the lowest percentage of S. In Table 6a, the relative fiber-type percent- presumes an association between specific ages for VL in two lemurs (2 and 3) have physiological phenomena and the morphobeen partitioned into superficialldeep and logical features examined herein. Conseproximalldistal regions. For all three fiber quently, the interpretation of the results is types, the disparity is greater between prox- subject to the “correctness” of these underimal and distal samples than between super- lying assumptions: (1)reduced physiological ficial and deep. However, what is provided cross-sectional area and calculated maxihere is actually a progressive series of sam- mum excursion are taken as estimates of the ples from proximal to distal and not a simple maximum effective force and maximum vesuperficialldeepdichotomy (Fig. 12,Table 6b). locity available to the animal from a muscle; An appreciably higher concentration of (2) all but a negligible portion of the free slower and more highly oxidative (S, FO) fi- elastic energy that contributes to the propulbers is found nearer to the attachment sites sive effort can be stored in the tendons rather 366 F.C. ANAPOL A N D W.L. JUNGERS TABLE 3. Fiber-type percentage cross-sectional (cross-sec) area; fiber cross-sect. area' Vastus medialis No. 1(n = 187) Mean fiber area (am') Range (pm2) % cross-sec. area No. 2 (n = 402) Mean fiber area (pm2) Range (pm') % cross-sec. area No. 3 (n = 444) Mean fiber area (pm2) Range (pm2) % cross-sec. area Composite lemur (n = 1,033) % cross-sec. area Range (%) Rectus femoris No. 1(n = 210) Mean fiber area (pm2) Range (pm2) % cross-sec. area No. 2 (n = 178) Mean fiber area (pm2) Range (pm') % cross-sec. area No. 3 (n = 237) Mean fiber area (pm2) Range (prn') % cross-sec. area Composite lemur (n : 625) % cross-sec. area Range (%) Vastus intermedius No. 1(n = 145) Mean fiber area (pm2) Range (pm') % cross-sec. area No. 2 (n = 221) Mean fiber area (pm') Range (pm2) % cross-sec. area No. 3 (n = 299) Mean fiber area (pm') Range (pm') % cross-sec. area Composite lemur (n = 665) % cross-sec. area Range (%) Vastus lateralis No. 1(n = 323) Mean fiber area (pm2) Range (pm') % cross-sec. area No. 2 (n = 693) Mean fiber area (pm2) Range (pm') % cross-sec. area No. 3 (n = 670) Mean fiber area (pm2) Range (pm') % cross-sec. area Composite lemur (n = 1,686) % cross-sec. area Ranee (%) I S FO F 1.378 (94111,524) 1.6 2,082 (975-2,837) 32.9 3.357 (3,236-3,937) 65.5 1,813 (1,314-2,317) 1.4 3,016 (1,196-4,147) 15.6 4,895 (3,462-6,391) 83.0 1,182 (787-1,330) 0.9 2,225 (1,385-3,391) 18.6 3,209 (1,658-4,014) 80.5 1.3 (0.9-1.6) 22.4 (15.6-32.9) 76.3 (65.5-83.0) 2,115 0.5 4,816 (2,594-9,009) 39.4 6,361 (6,014-7,164) 60.1 2,531 (2,517-2,550) 5.4 4,178 (3,264-5,111) 28.6 6,950 (6,806-7,140) 66.0 1,994 (1,803-2,719) 5.5 3,141 (2,168-4,743) 29.1 4,148 (3.655-6.002) 65.5 3.8 (0.5-5.5) 32.4 (28.6-39.4) 63.8 (60.1-66.0) 2,841 (2,633-3,711) 75.2 3,687 (2,633-4,479) 24.8 0.0 2,210 (1,721-2,552) 100.0 - - 0.0 0.0 2,948 (2,680-3,094) 71.9 3,482 (2,909-4,130) 28.1 0.0 0.0 0.0 82.4 (71.9-100.0) 17.6 (0.0-28.1) 0.0 2,020 (1,651-2,726) 13.8 2,180 (548-3,319) 28.7 4,151 (3,402-4,599) 57.5 2,123 (1,139-2,974) 8.6 3,969 (2,464-732) 21.8 5,657 (3,080-7,765) 69.7 1,794 (1,352-2,253) 5.2 3,124 (1,862-4,545) 37.0 3,575 (2,4354,655) 57.8 9.2 6.2-13.8) 29.2 (21.8-37.0) 61.7 (57.5-69.7) - 0.0 0.0 - IS, slow twitch - high oxidative; FO, fast twitch -high oxidative; F, fast twitch - low oxidative. 367 ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS 'r than in the contractile machinery of the muscle fibers (Alexander, 1974; Alexander and Bennett-Clark, 1977; but cf. Cavagna et al., 1980); and (3) low and high myosin ATPase activity are correlated to slow and fast rates of tension development and increasing concentrations of oxidative enzyme reflect increasing resistance to "fatigue" (DennyBrown, 1929; Burke et al., 1971). Electromyography of the quadriceps femoris in the locomotion and posture of Lemur fulvus is considered in detail elsewhere (Jungers et al., 1980). In general, all four muscles act synergistically to extend the crus during normal locomotor behaviors. However, vastus intermedius assumes almost exclusive responsibility in maintaining tension across the knee joint during posture. Because the four muscles insert via a common tendon closely applied to the anterior aspect of the knee, differences in moment arms among these muscles are presumed to be negligible. Consequently, gross mechanical differences related to leverage are virtually insignificant for the purposes of this analysis. I I - -0 I VASTUS LATERALIS Fig. 11. Mean percentage of cross-sectional area for S, FO, and F fibers in L. fuluus. Broken lines indicate range for three animals. TABLE 4. Twefactorial analysis of variance (ANOVAfimuscles versus animals (no replication))' Fiber Source of type variation S FO F ANOVA df MS SS FS A(musc1es) 3 7,739.31 2,579.77 2 9 . 46* B(anima1s) 2 211.80 105.90 1 . 2 1 N S Error 6 525.48 87.58 Total 11 8,476.59 A(musc1es) 3 346.33 115.44 1 . 77NS B(anima1s) 2 401.47 200.74 3 . 08NS Error 6 390.81 65.13 Total 11 1,138.61 A (muscles) 3 7,043.53 2,347.84 249. 77* B(anima1s) 2 63.25 31.62 3 . 36NS Error 6 56.40 9.4 Total 11 7,163.18 F,0512,3 = 5.14 F,0513,61= Fig. 12. Intramuscular variation in vastus lateralis (Lfuluus). Top-samples correspond to fiber-type percentages in Table 6a: 1,PxDp; 2, PxSu; 3, DsDp; 4, DsSu. Bottom: Numbered sections include fasciculi that correspond to samples in top picture (see Results and Table 6b). Su, superficial; Dp, deep; Px, proximal; Ds, distal. 4.76 'NS, not significant; *Significant difference (d = ,051;df, degrees of freedom; SS, sum of squares; MS, mean square; F,, sample statistics of F distribution. TABLE 5. Mann-Whitney U-test: all possiblepairs of muscles for each fiber type (a = .05) S Muscle pair VM x RF VM x VI VM x VL RF x VI RF x VL VI x VL ts 0.897 4.037 5.062 3.602 2.435 4.501 F FO Level of significance NS .05 .05 .05 .05 .05 ts Level of significance tS Level of significance 2.773 0.090 2.600 2.256 0.733 2.033 .05 NS .05 .05 NS .05 2.952 4.134 3.925 3.821 0.233 4.500 .05 .05 .05 .05 NS .05 F.C. ANAPOL AND W.L. JUNGERS 368 TABLE 6. Intramuscular variation of fiber types in V L for two lemurs (n = 1,363); Su: superficial; Dp: deep; P x wroximalt Ds: distal: M: mean (see Fie. 12) S Su Dp M a) Px Ds M Su FO Dp M Su F Dp M 7.2 7.9 7.6 30.7 32.5 31.6 62.2 59.6 60.9 7.0 5.4 6.2 29.8 24.6 27.2 63.2 70.3 66.6 7.1 6.7 30.3 28.5 62.7 64.8 b) PxDp (1) PxSuQ) DsDd3) DsSu(4) S FO F 7.9 7.2 5.4 7.0 32.5 30.7 24.6 29.8 59.6 62.2 70.3 63.2 TABLE Z Summary of morphological data for Lemur fulvus-ratios; for each muscle, its derived value (see Table 3) is divided by the lowest value o f the four VM RF If 4.7 Reduced physiological cross-sec. 1.0 area (% of group) h 4.4 MassPo 4.5 1.0 ttflf + tt Fiber types (% cross-sec. area) S 1.0 FO 1.3 F 1.2 VI VL 1.0 3.6 1.1 1.5 2.6 10.1 1.0 1.0 2.6 1.1 1.1 2.4 1.5 1.5 2.4 2.9 63.4 1.8 1.0 1+ - 7.1 1.7 1.0 A profile for each muscle of the quadriceps femoris in the brown lemur is presented below. The criteria upon which these profiles are based are included only in the characterization of vastus medialis but are the same for rectus femoris, v. intermedius, and v. lateralis. In order to facilitate comparisons among muscles, parametric data are expressed as ratios in Table 7: for every parameter, the value of each muscle is divided by the lowest of the fmr. Vastus medialis (1) Least potential effective tetanic force (estimated by reduced physiological crosssectional area); (2) highest potential velocity (estimated by calculated maximum excursion and massPo); (3) least potential conservation of energy (estimated by length of tendon per fasciculus); and (4)least time to maximum recruitment as implied by a relatively high percentage crosssectional area of fast twitch, low oxidative (F) muscle fibers. Coupling long fasciculus excursion with a preponderance of fibers that quickly rise to peak tension, VM may be specialized to augment the most rapid or dynamic locomotor activities with a late-onset surge of additional force. With a relative dearth of oxidative fibers, VM is poorly suited for sustained contraction and, accordingly, would not need to conserve energy (which is believed to be required by “endurance-type” behaviors). The low proportion of tendon per fasciculus implies that this muscle is probably sacrificing some energy (that would otherwise be conserved by transmitting force to bone through noncontractile tendon) in favor of increasing its potential velocity by the inclusion of contractile material along almost the full length of the whole muscle. The fast-fibered VM may be saved for more strenuous activities which require rapid recruitment of additional muscle to augment the propulsive effort. This would be compatible with current theory of motor-unit recruitment which hypothesizes that primarily Sand FO-type motor units are recruited for activities as strenuous as fast running and that F-type fibers are probably reserved for more vigorous locomotor activities (Walmsley et al., 1978; Armstrong, 1980). Rectus femoris (1) Moderate potential effective tetanic force; (2) lowest potential velocity; (3) highest potential conservation of energy; and (4)low time to maximum recruitment with high aerobic capacity. The morphological features of RF suggest that this muscle has the potential to sustain either isometric or eccentric contraction during normal locomotion. Firstly, although RF is the sole leg extensor to cross both the hip and knee joints, its estimated maximal excursion is surprisingly small. Secondly, the bipinnate arrangement of its muscle fibers (contra Junders et al., 1980) allows a relatively high force output per gram of mass. Thirdly, the high percentage of FO fibers would allow muscle stiffening to occur almost instantaneously and repeatedly for extended periods of time. ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS Theoretically, RF could behave as a n elastic body, storing strain energy when stretched during either hip extension andor knee flexion. This energy would be released when the hip is flexed or the knee is extended. Not only would coordination between the hip and knee joints continually be maintained, but the extensors of the thigh (e.g., femorococcygeus) could effectively contribute to leg extension through a stiffened, two-joint RF? Of interest is that RF is recruited twice during a single-step cycle, both as a n extensor of the leg during support and again during swing phase as it flexes the hip in order to recover the limb for the next step cycle (Jungers et al., 1980). A morphological configuration which allows rapid development and continuous maintainance of isometric tension seems well suited for a multifunctional muscle that flexes one joint and extends another. 369 percentage of highly oxidative fast-contracting fibers in a postural muscle would allow a t least a portion of the muscle to keep pace with the more rapidly contracting VM, RF, and VL during relatively vigorous locomotor activities such as galloping and leaping (a constituative locomotor activity in lemur). Vastus lateralis (1)Highest potential effective tetanic force; (2) second-highest potential velocity; (3) relatively high potential conservation of energy; and (4)most heterogeneous distribution of fiber types. Among the quadriceps, VL is undoubtedly the primary contributor to locomotion in L fuluus. Its extraordinary mass includes statistically significant proportions of all three classes of fibers, thereby broadening its range of recruitment. This muscle, which obtains Vastus intermedius almost 60% of the combined mass of the (1)Substantial potential effective tetanic group, appears to optimize a combination of force, velocity, economy, and endurance. force; The architectural configuration observed in (2) low potential velocity; (3) relatively high potential conservation of V1 is especially suitable for leaping behavior. Alezais (1900) and Stern (1971) concluded energy; and (4)longest time to maximum recruitment that the small proximal attachments of the vasti peripherales in leaping rodents and priwith highest aerobic capacity. VI is especially well suited for postural be- mates allow the muscle fibers to be longer havior and is also recruited at high levels than those in nonleaping or climbing genera during all progressive activities (Jungers et in which the corresponding attachments are al., 1980). It is the deepest of the quadriceps, more extensive. In L. fuluus, a n exceptionwith muscle fibers arising directly from bone ally agile leaper, VM and VL both arise from and passing close to the knee joint. All fea- circumscribed attachments at approximately tures of its morphology suggest the ability to the same transverse plane at the proximal maintain substantial isometric tension as end of the femur. Indeed, both are composed well as slow, consistent, isotonic phasic activ- of relatively long fasciculi. However, while the fibers of VM are parality for extensive periods of time. This is indicated by a population of entirely high- lel to the presumed line of action of the whole oxidative muscle fibers, most of which ex- muscle, the fibers of VL are unipinnately hibit myosin ATPase reactivity similar to arranged [in emendation of the long parallel that known for slowly contracting fibers fibers previously described by Jungers et al. (Denny-Brown, 1929; Henneman and Olson, (198011. The mean fasciculus length is slightly 1965; Burke et al., 1971; McDonagh et al., less than one-third of that for VM, despite 1980). This fiber-type composition is typical their similarity in whole muscle length. With relatively long pinnate fibers, VL is of demonstrably “postural” muscles, i.e, soleus (6.Smith et al., 1977; Walmsley et al., able to implement both high velocity and 1978; Burke, 1980; Smith et al., 1980; Gardi- considerable tension. Firstly, pinnation allows a n increase in the number of muscle ner et al., 1982). Substantial complements (24.8% and 28.1%) of FO fibers are present in VI in two specimens (Table 3). This may reflect a n ex‘Cf. Walmsley et al. (1978).who suggested that the inherent tended range of motor-unit recruitment in a n stiffness in the fully active medial gastrocnemius in cat would otherwise slow muscle (see Walmsley et al., transmit forces from the powerful knee extensors to the 1978, for further discussion). A respectable calcaneus. 370 F.C. ANAPOL AND W.L. JUNGERS fibers (i.e., the number of cross-bridges in parallel) and, therefore, the amount of force that can be applied to the tendon of insertion. Secondly, the excursion of a pinnate whole muscle is not merely proportional to but longer than that of its constituent fibers (see Gans, 1982; Muhl, 1982). Thirdly, because they contain more sarcomeres in series, long fibers (pinnate or parallel) are able to generate a n equivalent excursion with less shortening per sarcomere (Muhl, 1982). Consequently, longer fibers can maximize the length-tension relationship by remaining within a range of greater thirdthick filament overlap (see Goslow et al., 1977; Walmsley and Proske, 1981; Muhl, 1982). Because the fasciculi of VL are considerably longer than those of either RF or VI, it is the muscle best suited for production of force and the pinnate muscle best suited for velocity. With a significant percentage of cross-sectional area composed of type S fibers, the fiber-type distribution of VL is the most heterogeneous of the four muscles. Although predominated by fast fibers (FO plus F) both RF and VL are less specialized than VM with its significantly higher percentage of F fibers. Because of their relative heterogeneity, RF and VL would be expected to be active over a wider range of locomotor activities than either VM or VI. Of additional interest is the distribution of fiber types within VL. In extensor muscles, the percentage of slow fibers is thought to decrease from deep to superficial, presenting a more-or-less-“stratified” arrangement (Armstrong, 1980).This tenet could not apply to unipinnate muscles (such as VL), whose fibers are oriented superficial to deep, because the enzymatic activities within individual fibers are homogeneous throughout their lengths (Pette et al., 1980). Alternatively, in VL both S and FO fiber types were most highly concentrated near the joints (Table 6b). The fastest fibers with the lowest oxidative enzyme levels were most highly concentrated at midbelly. This is comparable to results reported in lizard, in which FOG and tonic fibers (sensu stricto) predominate around joints in the hindlimb muscles with FG fibers composing most of the overall muscle mass (Putnam et al. 1980). Thus, the quadriceps femoris in L. fuZuus is composed of four morphologically and (probably) functionally distinct muscles. Each is characterized by a constellation of physiologically related architectural and histochemi- cal features that underlies its particular contribution to the propulsive effort. One interpretation is that this broadens the range of functional capabilities of the quadriceps as a group. If in all of the four muscles the architecturally related physiological alternatives (e.g., force versus velocity) were compromised, full realization of any parameter would be limited for the entire group. However, the brown lemur retains three relatively specialized muscles: (1) VM-high velocity at the expense of force and energy; (2) RF-moderately high force and low energy cost a t the expense of velocity; and (3) VI-a postural muscle. By contrast, there is only a single ‘‘ comprehensive” muscle (VL). One might suspect that the notably complex positional behavior of primates should be associated with equally complex anatomical and physiological substrates. Studies of primate locomotor behavior that tend to treat species as if each were “specialized” for one particular locomotor mode have been soundly criticized by several workers (e.g., Stern and Oxnard, 1973; Mittermeir and Fleagle, 1976; Ripley, 1977) because this practice ignores both the variation in complexity among locomotor “niches” (Stern and Oxnard, 1973) and the requisite neurological and myological complexity in the animals, i.e., the dynamic input to complex locomotor behaviors. Any hypothetical functional division of labor within lemur quadriceps may very well be associated with a complex neuromuscular requirement for rapidly traversing a threedimensional arboreal habitat. Although retaining the ability to move quite rapidly on the ground with ease, the quadrupedal brown lemur spends at least 90% of its time walking, running, and leaping through the upper canopy levels of the forest (Walker, 1967; Ward and Sussman, 1979; after Sussman, 1972).They (variably) prefer sloping and horizontal supports of large diameter and have been observed to leap up to 4 m between adjacent branch systems in lieu of continuous passages (Tattersall, 1982). When compared to other mammals (e.g., cats), lemurs appear to be relatively flexible with respect to neural programming of interlimb coordination (Jungers and Anapol, 1985);i.e., a single motor “program” will not suffice for all gaits. This implies that lemurs rely more upon facultative peripheral input to modulate their neuromuscular outputs. The anatomical substrates of their locomotor behavior would therefore be expected to sub- ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS serve and reflect this requisite flexibility. In essence, the great complexity of an arboreal habitat (e.g., variability among substrate diameters and juxtaposition of trees and branches) probably requires exceptional precision with respect to control of the locomotor apparatus; the neuromuscular system must be able to provide rapid adaptability to a grossly irregular and often precarious arboreal topography. The expression of the standard deviation as a percentage of the mean of the population can be used to compare populations of appreciably differing means (Sokal and Rohlf, 1981). Thus, the morphological diversity among VM, RF, VI, and VL was further illustrated by calculating the coefficient of variation adjusted for small sample bias (V*) for each parameter in this study (Table 8). The V* values indicate that for most of these parameters, all muscles vary considerably about the mean of the four. Of all the parameters included in Table 8, the percentage cross-sectional area of FO fibers has the lowest coefficient of variation among muscles. This class includes a wider variety of cells than either the S or F classes. It includes cells of both intermediate and high myosin-ATPase reactivity, the former having been correlated to slightly slower speeds of contraction (Burke, 1981). Fast twitch-high oxidative fibers also include two levels (dark and medium) of concentration of oxidative enzymes, thereby providing additional variety in aerobic capabilities (socalled “resistance to fatigue”). The range of cross-sectional areas of FO fibers is also considerably broader than that of either S or F classes (Table 3). In addition, among the three classes of motor units in cat gastrocnemius, the FO units have the fewest muscle fibers per axon (Burke and Tsairis, 1973). If this relationship holds for lemurs, TABLE 8. Coefficients of uariation’ Parameter lf Reduced physiological cross-see.area h MassPo ttfif + tt Fiber types S FO F v* * s*, 89.5 98.1 85.4 85.7 37.9 k 31.6 k 34.7 + 30.2 k 30.3 k 13.4 171.2 k 60.5 27.9 k 9.9 72.1 25.5 + ‘V*, coefficient of variation corrected for small sample size: sV*, standard error of the statistic. 371 the FO portion of cross-sectional area could be recruited in variably smaller parcels than either the F or S units. Despite the occurrence of dramatic architectural differences among the four heads of the quadriceps in the brown lemur, a means is provided by which the different potentials within and among muscles can be realized to produce a smooth and more-or-less-continuous power band. In other words, while disparate architectural profiles expand the range of functional capabilities for the quadriceps as a group, physiological continuity is maintained among muscles by the presence of substantial pools of an intermediate, relatively heterogeneous class of fibers that serves as a “common denominator.” Other mammals whose habitats appear to require flexible circuitry for neuromuscular control also exhibit a relatively balanced fiber population. For example, observations of Tupaia glis indicate that this species is both arboreal and terrestrial. However, the arborealherrestrial dichotomy may be artificial for an animal this size because the relatively cluttered forest floor offers the same impediments to locomotory progression as encountered in the trees (Jenkins, 1974). Thus, the locomotor repertoire of tree shrews is adaptive to “two major topographical features” which favor a “highly flexible and versatile locomotor pattern”: (1) surfaces that are not level, and (2) a spatial arrangement of locomotor surfaces that is disordered (Jenkins, 1974, pp. 110-111). The critical factors linking lemur and tree shrew locomotor behavior are (1)their habitual exploitation of a three-dimensional environment regardless of the extent to which it is “arboreal” or “terrestrial”, and (2) such exploitation often executed at high velocities (as in running, galloping, or leaping). The quadriceps in both lemur and tree shrew are composed of relatively heterogeneous (no category less than 20%) distributions of fiber types, which includes a large percentage of FO fibers (Table 9). The association between the behavioral and histochemical features characteristic of the lemur and tree shrew can be contrasted to that for cat (Felis domesticus), dog (Canis familiaris), loris (Nycticebus coucang), and lesser bushbaby (Galago senegalensis). For these taxa, percentages (unadjusted for crosssectional area) of three major classes of fibers in quadriceps femoris are presented in Table 9. 372 F.C. ANAPOL AND W.L. JUNGERS TABLE 9. Fiber-type percentages in quadriceps femoris for various species (decreasing quantities of S fibers) Loris (Sickles and Pinkstaff, 1981) VM RF VI VL Mean quadriceps femoris Dog (Armstrong e t al., 1982) VM RF VI VL Mean quadriceps femoris Cat (Ariano e t al., 1973) VM RF VI VL Mean quadriceps femoris Lemur (Anapol, 1984) VM RF VI VL Mean quadriceps femoris Tree shrew (Sickles and Pinkstaff, 1981) VM RF VI VL Mean quadriceps femoris Lesser bushbaby (Sickles and Pinkstaff,l981) VM RF VI VL Mean quadriceps femoris Cats and dogs are the most cursorially adapted of the carnivores (Howell, 1944).Felids, in general, are best characterized by a “prey-capture” mode of behavior (Eisenberg, 1981). In domestic cats this entails (1)a fast run in a crouched position or slow stalking of prey, and (2) springing onto prey during which the cat “darts forward flattened against the ground, either running or with several bounds” (Leyhausen, 1979, p. 6; also see Gambarayan, 1974). They usually do not chase their prey among the branches nor among the trees, and their large size effectively eliminates a considerable portion of the impediments to locomotion (such as uneven terrain) that faces animals the size of tree shrews. The fiber-type distribution in quadriceps femoris in cats is relatively polar- s PO 64 55 86 52 64 36 45 14 48 36 - 39 41 88 43 53 61 59 12 57 47 - 13 22 98 27 40 18 69 61 3 8 85 15 28 27 38 15 33 29 70 54 52 44 1 10 92 75 57 25 32 8 - 4 27 72 53 23 20 1 3 49 24 45 51 14 34 75 52 86 53 - 13 ~ 17 2 17 14 F - - - - 56 47 ized (Table 9). Preferential selection of slow and fast-low oxidative fibers (at the expense of FO) would seem to be adaptive for both their stalking behavior (which is executed in poised position, regardless of speed) and springing (pouncing) behavior, respectively. By contrast, dogs pursue prey terrestrially for relatively long distances at moderate running speeds (Howell, 1944; Gambarayan, 1974; Eisenberg, 1981). Of no surprise is the exclusive occurrence of highly oxidative fibers (S; FO) in all of their hindlimb muscles (Armstrong et al., 1982). Dogs (and loris) lack F fibers and have the least-heterogeneous fiber-type distributions of the animals presented here. They both exhibit fiber profiles which emphasize slower contraction and high endurance. ARCHITECTURE AND HISTOCHEMISTRY IN LEMUR QUADRICEPS FEMORIS The slow climbing of Nycticebus coucang (Walker, 1967) presents a relatively less precarious locomotory situation for this arboreal quadruped. With usually two or three feet grasping the support (Walker, 1967; Dykyj, 1980), the need for rapid adjustment to the variety of substrates found among the trees is dramatically reduced. Constituent fiber types in loris are exclusively the high-endurance variety, most of which are slow contracting (Sickles and Pinkstaff, 1981). Electromyography has indicated that some leg muscles remain active for longer periods of time in slow loris than in more rapidly moving animals covering the same distance (see Anapol and Jungers, 1982; Jungers et al., 1983). At the other extreme (relatively few S fibers) is Galago senegalensis. Lesser busbabies are primarily tree dwellers that leap among branches with great agility, although a t slower speeds they have been observed to move quadrupedally and have also been observed “waddling” quadrupedally on the ground (Walker, 1967). Among the four arboreal (or semiarboreal, in the case of tree shrews) species presented in Table 9, increasing populations of faster and less oxidative fibers (loris < tree shrew < lemur < bushbaby) are correlated with a n increasing proclivity for leaping a s a primary mode of locomotion (see Gillespie et al., 1974; Sickles and Pinkstaff, 1981); for example, leaping is a much-higher-frequency mode in Galago than in L fulvus (Walker, 1967; after Attenborough, 1961). Moreover, results from in vivo mechanical studies of hindlimb muscle function in G. senegalensis predict the presence of a high proportion of fast twitch-low oxidative fibers in the quadriceps femoris (Hall-Craggs, 1974). The results presented here encourage speculation regarding expected intramuscular morphological configurations in other primates. Great architectural disparity among synergists coupled with overall fiber type heterogeneity (including a substantial proportion of fast twitch-high oxidative fibers) is the likely set of character states in animals whose locomotor repertoires require rapid adaptability to complex substrates. Among anthropoid primates, this would include those taxa which exploit arboreal environments with rapid quadrupedalism and leaping among trees and branches, e.g., Saimiri or Aotus (Fleagle and Mittermeier, 1980; 373 Plaghki et al., 1981) and leaf monkeys such as Presbytis melalophos (Fleagle, 1977). Alternatively, hindlimb muscles in primates committed to locomotor modes less dependent upon such flexibility, e.g., Erythrocebus (terrestrial quadrupedalism), Tarsius (obligatory leaping), or Homo (bipedality), would exhibit less architectural diversity and more polarization of fiber types (predominantly slow and fast twitch-low oxidative fibers). Species whose locomotor repertoires habitually include more than a single locomotor mode, e.g., L catta, which spends considerably more of its time on the ground than other lemur species (Walker, 1967; after Jolly, 1966; Ward and Sussman, 19791, present different possibilities. For example, would the muscle morphology of these forms adhere to a model that emphasizes their most frequent behavior mode or one that lies somewhere between the extremes described above? Nevertheless, the results reported herein support the hypothesis that the behavior of a n animal is associated intimately with the architectural and histochemical bases of its anatomy. Whether the interpretation of these results can be extended without modification to other vertebrate groups and to other musculoskeletal systems in general remains to be tested. ACKNOWLEDGMENTS We wish to express our sincere appreciation to David S. Carlson, Maynard M. Dewey, John G. Fleagle, Susan W. Herring, Farish A. Jenkins, Jr., Jack T. Stern, Jr., and anonymous reviewers for their invaluable comments and criticisms on the manuscript. A special thank you to Luci Betti for preparing the expert graphs and illustrations, and to George Boykin, Norman Creel, Steve Hartman, Eileen Schneider, and Robert Skinner for their technical advice and assistance. Financial support was provided by the Department of Anatomical Sciences, State University of New York a t Stony Brook, NIH biomedical research grant BSRG 2SO7RRO573610 to W.L. Jungers, and NSF research grant BNS 8119664 to J . T. Stern, Jr., W.L. Jungers, and R.L. Susman. LITERATURE CITED Alexander, RMcN, (1974) The mechanics of jumping by a dog (Canis familiaris).J. Zool., Land. 173:549-573. Alexander, RMcN, and Bennet-Clark, HC (1977) Storage of elastic strain energy in muscle and other tissues. Nature 265114-117. 374 F.C. ANAPOL AND W.L. JUNGERS Alezais, H (1900) Le quadriceps femoral des sauteurs. C. Eisenberg, JF (1981) The mammalian radiations. Chicago: Univ. Chicago Press. R. Soc. Biol. 52:510-511. Anapol, FC (1984) Morphological and Functional Diver- Fleagle, JG (1977) Locomotor behavior and muscular sity Within the Quadriceps Femoris in Lemur fuluus: anatomy of sympatric Malaysian leaf monkeys (PresArchitectural, Histochemical, and Electromyographic bytis obscura and Presbytus melulophos). Am. J. Phys. Considerations. Ph.D. Dissertation. State University Anthropol. 46:297-308. of New York at Stony Brook. Fleagle, JG, and Mittermeier, RA (1980)Locomotor behavior, body size, and comparative ecology of seven Anapol, FC, and Jungers, WL (1982)Division of labor in muscles: A functional comparison of medial gastrocneSurinam monkeys. Am. J. Phys. Anthropol. 52301mius and soleus in lemur and loris. Am. J. Phys. An314. thropol. 57:166-167 (Abstract). Gambarayan, P P (1974) How Animals Run. New York: Ariano, MA, Armstrong, RB, and Edgerton, VR (1973) John Wiley and Sons. Hindlimb muscle fiber populations of five mammals. Gans, C (1982) Fiber architecture and muscle function. J. Histochem. Cytochem. 215-55. Exerc. Sport. Sci. Rev. 10:160-207. Armstrong, RB (1980)Properties and distributions of the Gans, C, and Bock, WJ (1965) IV.The functional signififiber types in the locomotory muscles of mammals. In cance of muscle architecture-a theoretical analysis. K Schmidt-Nielsen,L Bolis, and CR Taylor (eds): ComErgeb. Anat. Entwickl. 38:115-142. parative Physiology: Primitive Mammals. Cambridge: Gardiner, KR, Gardiner, PF, and Edgerton, VR (1982) Cambridge Univ. Press, pp. 243-254. Guinea pig soleus and gastrocnemius electromyoArmstrong, RB, Marum, P, Saubert, IV, CW, Seehergrams at varying speeds, grades, and loads. J. Appl. man, HJ, and Taylor, CR (1977) Muscle fiber activity Physiol. Respir. Environ. Exerc. Physiol. 52451-457. as a function of speed and gait. J. Appl. Physiol. Respir. Gillespie, CA, Simpson, DR, and Edgerton, VR (1974) Environ. Exercise Physiol. 43:672-677. Motor unit recruitment as reflected by muscle fiber Armstrong, RB, Saubert, IV, CW, Seeherman, HJ, and glycogen loss in a prosimian (bushbaby) after running Taylor, CR (1982) Distribution of fiber types in locoand jumping. J. Neurol. Neurosurg. Psychiatr. 37:817motory muscles of dogs. Am. J. Anat. 163237-98. 824. Ashton, EH, and Oxnard, CE (1963) The musculature of Goslow, GE, Jr., Cameron, WE, and Stuart, DG (1977) the primate shoulder. Trans. Zool. SOC. Lond., 29:553Ankle flexor muscles in the cat: Length-active tension 650. and muscle unit properties as related to locomotion. J . Attenborough, D (1961) Zoo quest to Madagascar. LonMorphol. 153:23-28. don: Luttenvorth. Guth, L, and Samaha, FJ (1970)Procedure for the histoBenninghoff, A,, and Rollhauser, H (1952) Zur inneren chemical demonstration of actomyosin ATPase. Exp. Neurol. 28:365-367. Mechanik des gefiederten Muskels. Pflugers Arch. ges. Physiol. 254:527-548. Hall-Craggs, ECB (1974) Physiological and histochemiBiewener, A, Alexander, RMcN, and Heglund, NC (1981) cal parameters in comparative locomotor studies. In Elastic energy storage in the hopping of kangaroo rats RD Martin, GA Doyle, and AC Walker (eds): Prosimian (Dipodomys spectubilis). J. Zool., Lond. 195:369-383. Biology. Pittsburgh: Univ. Pittsburgh Press, pp. 829845. Brooke, MH, and Kaiser, KK (1970) Muscle fiber types: How many and what kind? Arch. Neurol. 23:369-379. Haxton, HA (1944) Absolute muscle force in the ankle flexors of man. J. Physiol. (Lond.) 103:267-273. Burke, RE (1975) A comment on the existence of motor unit Yypes.” In DB Tower (ed): The Basic Neurosci- Henneman, E., and Olson, CB (1965)Relations between ences, Vol. 1: The Nervous System. New York: Raven structure and function in the design of skeletal musPress, pp. 611-619. cles. J. Neurophysiol. 28581-598. Burke, RE (1980) Motor unit types: Functional speciali- Howell, AB (1944) Speed in Animals. Chicago: Univ. zation in motor control. TINS 3:255-258. Chicago Press. Burke, RE (1981)Motor units: Anatomy, physiology, and Jenkins, FA, Jr., (1974) Tree shrew locomotion and the functional organization. In VB Brooks (ed): Handbook origins of primate arborealism. In FA Jenkins, Jr. (ed): of Physiology, Sec. 1: The Nervous System, Vol. 11, Part Primate Locomotion. New York: Academic Press, pp. 1.Bethesda: Am. Physiol. SOC.,pp. 345-422. 85-115. Burke, RE, Levine, DN, Zajac, 111, FE, Tsairis, P, and Jolly, A (1966) Lemur Behavior: A Madagascar Field Engel, WK (1971)Mammalian motor units: PhysiologStudy. Chicago: Univ. Chicago Press. ical-histochemical correlation in three types in cat gas- Jouffroy, FK (1962) La musculature des membres chez trocnemius. Science 174:709-712. les lkmuriens de Madagascar. Etude descriptive et Burke, RE, and Tsairis, P (1973) Anatomy and innervacomparative. Mammalia Z6(suppL 2):l-326. tion ratios in motor units of cat gastrocnemius. J. Jungers, WL, and Anapol, FC (1985) Interlimb coordiPhysiol. Gond.) 234:749-765. nation and gait in the brown lemur (Lemur fuluus) and Cavagna, GA, Citterio, G, and Jacini, P (1980) Elastic the talapoin monkey (Miopithecus tulapoin). Am. J. storage: Role of tendons and muscles. In K SchmidtPhys. Anthropol. 6789-97. Nielsen, L Bolis, CR Taylor (eds): Comparative Physi- Jungers, WL, Joufhoy, FK, and Stern, Jr., JT (1980) ology: Primitive Mammals. Cambridge: Cambridge Gross structure and function of the quadriceps femoris Univ. Press, pp. 231-242. in Lemur fuuluus: An analysis based on telemetered Denny-Brown, DE (1929) The histological features of electromyography. J. Morphol. 164:287-299. striped muscle in relation to its functional identity. Jungers, WL, Stern, Jr., Sr, and Jouffroy, FK (1983) Roc. R. Soc. Lond. (Biol.) 104:371-411. Functional morphology of the quadriceps femoris in Dubowitz, V, and Brooke, MH (1973) Muscle biopsy: A primates: A comparative anatomical and experimental modern approach. London: WB Saunders. analysis. Annales des Sciences Naturelles, Zoologie (Paris) 5:lOl-116. Dykyj, D (1980)Locomotion of the slow loris in a designed substrate context. Am. J. Phys. Anthropol. Leyhausen, P. (1979) Cat Behavior. New York: Garland 52:577-586. STPM Press. ARCHITECTURE AND HISTOCHEMISTRYIN LEMUR QUADRICEPSFEMORIS McDonagh, JC, Binder, MD, Reinking, RM, and Stuart, DG (1980) Tetrapartite classification of motor units of cat tibialis posterior. J. Neurophysiol. 44:696-712. Mittermeier, RA, and Fleagle, J G (1976) The locomotor and postural repertoires of Ateks geoffroyi and Colobus guerza, and a reevaluation of the locomotor category semibrachiation. Am. J. Phys. Anthropol. 45(2):235255. Muhl, Z (1982) Active length-tension relation and the effect of muscle pinnation on fiber lengthening. J. Morphol. 173:285-292. Murie, J, and Mivart, StG (1872)On the anatomy of the Lemuroidea. Trans. Zool. Soc. Lond. 71-114. Murphy, RA, and Beardsley, AC (1974)Mechanical properties of the cat soleus muscle in situ. Am. J. Physiol. 2271008-1013. Nemeth, P, and Pette, D (1981)Succinate dehydrogenase activity in fibres classified by myosin ATPase in three hind limb muscles of rat. J. Physiol. 320:73-80. Peter, JB, Barnard, RJ, Edgerton, VR, Gillespie, CA, and Stempel, KE (1972) Metabolic profiles of three fiber types of skeletal muscle in Guinea pigs and rabbits. Biochemistry 11:2627-2633. Pette, D, Wimmer, M, and Nemeth, P (1980) Do enzyme activities vary along muscle fibers? Histochemistry 67225-231. Petter, JJ, Albignac, R, and Rumpler, Y (1977) MammiEres Lemuriens (Primates, F’rosimiens). Faune de Madagascar 44:l-513. Plaghki, L. Goffart, M, Beckers-Bleukx, G, and Moureau-Lebbe, A (1981) Some characteristics of the hind limb muscles in the leaping night monkey Aotus triuirgatus (Primates, Anthropoidae, Cebidae). Comp. Biochem. Physiol. 70A:341-349. Putnam, RW, Gleeson, lT, and Bennett, AF (1980) Histochemical determination of the fiber composition of locomotory muscles in a lizard, Dipsosaurus dorsalis. J. Exp. Zool. 214:303-309. Ripley, S (1977) Gray zones and gray langurs: Is the “semi” concept seminal? Yrbk. Phys. Anthropol. 20376-394. Sacks. RD. and Rov. RR (19821 Architecture of the hind limb muscles of cats: Functional significance. J. Morphol. 173:185-195. Schumacher, GH (1961) Funktionelle Morphologie der Kaumuskulatur. Gustav Fischer, Jena. (Trans. by Z. Muhl). Sickles, DW, and Pinkstaff, CA (1981) Comparative histochemical study of prosimian primate hindlimb muscles. 11. Populations of fiber types. Am. J. Anat. 160:187-194. Smith, JL, Betts, B, Edgerton, VR, and Zernicke, RF 375 (1980) Rapid ankle extension during paw shakes: Selective recruitment of fast ankle extensors. J. Neurophysiol. 43:612-620. Smith, JL, Edgerton, VR, Betts, B, and Collatos, TC (1977) EMG of slow and fast ankle extensors of cat during posture, locomotion, and jumping. J. Neurophysioi. 40:503-513. Sokal, RR, and Rohlf, FJ (1981) Biometry, 2nd ed. San Francisco: WH Freeman. Spector, SA, Gardiner, PF, Zernicke, RF, Roy, RR, and Edgerton, VR (1980) Muscle architecture and forcevelocity characteristics of cat soleus and medial gastrocnemius: Implications for motor control. J. Neurophysiol. 44:951-960. Stein, JM, and Padykula, HA (1962)Histochemical classification of individual skeletal muscle fibers of the rat. Am. J. Anat. 110(2):103-124. Stern, JT, Jr. (1971) Functional myology of the hip and thigh of cebid monkeys and its implications for the evolution of erect posture. Bibl. Primatol. 14:l-319. Stern, JT,Jr., and Oxnard, CE (1973) Primate locomotion: Some links with evolution and morphology. Primatologia 4(11):1-92. Sullivan, TE, and Armstrong, RB (1978)Rat locomotory muscle fiber activity during trotting and galloping. J. Appl. Physiol. Respir. Environ. Exercise Physiol. 44:358-363. Sussman, RW (1972) An Ecological Study of Two Madagascan Primates: Lemur fulvus rufus (Audeburt) and Lemur cattu (Linnaeus). Ph.D. Dissert. Duke University. Tattersall, I(1982) The Primates of Madagascar. New York Columbia Univ. Press. Taylor, CR (1978) Why change gaits? Recruitment of muscle and muscle fibers as a function of speed and gait. Am. 2001. 18:153-161. Walker, A (1967) Locomotor Adaptations in Recent and Fossil Madagascar Lemurs. Ph.D. Thesis. Univ. of London. Walmsley, B, Hodgson, JA, and Burke, RE (1978)Forces produced by medial gastrocnemius and soleus muscles during locomotion in freely moving cats. J. Neurophysiol. 41:1203-1216. Walmsley, B, and Proske, U (1981) Comparison of stiffness of soleus and medial gastrocnemius muscles in cats. J. Neurophysiol. 46:250-259. Ward, SC, and Sussman, RW (1979)Correlates between locomotor anatomy and behavior in two sympatric species of Lemur. Am. J. Phys. Anthropol. 50:575-590. Weber, EF (1851) h e r die Langenverhaltnisse der Fleischfasern der Muskeln im allgemeinen. Ber. K. sachs. Ges. Wiss. nat. phys. Kl:64-86.