Effects of ontogeny and sexual dimorphism on scapula morphology in the mountain gorilla (Gorilla gorilla beringel).код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 98:431445 (1995) Effects of Ontogeny and Sexual Dimorphism on Scapula Morphology in the Mountain Gorilla (Gorilla gorilla beringel) ANDREA B. TAYL,OR School of Physical Therapy, Sltppery Rock University, Slippery Rock, Pennsylvania 16057 KEY WORDS Mountain gorilla, Scapula, Ontogeny, Allometry Locomotion, Sexual Dimorphism ABSTRACT Scapular measurements were obtained from growth series of the sexually dimorphic mountain gorilla (Gorilla gorilla beringei). Juveniles, subadults, and adults were compared to determine if scapula morphology varies with age. Analyses reveal significant (P < 0.05) differences in scapula form for shape ratios of length vs. breadth, length vs. infraspinous fossa length, and length vs. spine length. Males and females were also compared to determine if sexual dimorphism in scapula morphology is a consequence of differential extension of common patterns of relative growth. Analyses reveal that scapula proportions are ontogenetically scaled. Data indicate t h a t male scapulae grow a t a faster rate and for a longer duration than females. Results of comparisons of males and females suggest that unique adaptations to different ecological niches have not evolved between the sexes despite sexual differences in frequency of patterns of locomotor behavior. By contrast, age-related variation in scapula morphology may be linked to differences in locomotor behavior during ontogeny. 0 1995 Wiley-Liss, Inc Gorillas are the largest and most sexually dimorphic of the extant hominoid primates. Males weigh approximately 200kg, and females average around 90kg (Fleagle, 1988). I n addition to being sexually dimorphic in body size, field studies of the mountain gorilla (G.g. beringei) have documented sexual differences in frequency of locomotor behavior between adults and patterns of variation in locomotion as a function of agelsize (Schaller, 1963; Tuttle and Watts, 1985). Although there may be sexual and ontogenetic differences in skeletal morphology a s a function of locomotor behavior in mountain gorillas, differences between sexes in adult morphology are best understood in the context of ontogenetic data since divergence between sexes can be achieved in numerous ways. Therefore, patterns of size variation are often explored allometrically utilizing ontogenetic series because such analyses provide insights into processes of growth and devel0 1995 WILEY-LISS, INC. opment which underlie differences in size. Indeed, many analyses of adult postcranial morphology have revealed that sex differences in adult form are really the products of differential extension of common patterns of relative growth (Shea, 1981, 1984, 1992; Jungers and Susman, 1984; Inouye, 1992; Jungers and Cole, 1992; Ravosa et al., 1993). Behavioral studies firmly establish the dominant mode of locomotion in gorillas to be a specific form of quadrupedalism known as terrestrial knuckle-walking (Yerkes and Yerkes, 1929; Schaller, 1963, 1965; Tuttle and Watts, 1985). Although not unique among large-bodied hominoids, gorillas and chimpanzees share a peculiar form of knuckle-walking that differs from that of the orang-utans. In particular, the African apes Received June 20, 1994; accepted June 2, 1995. Address reprint requests to Andrea B. Taylor, School of Physi- cal Therapy, Slippery Rock University, Slippery Rock, PA 16057. 432 A.B. TAYLOR share a suite of morphological features which includes marked shortening of the long digital flexor tendons, tightly packed articular joint surfaces a t the metacarpophalangeal joints, extremely strong ligaments and tendons in the hand, and specialized pads over the knuckles (Tuttle, 1967). Together, these features provide for special weight-bearing adaptations on the dorsal aspects of the middle phalanges of the flexed hand digits II-IV which enable the animal to locomote via a combination of digitigrade hands and plantigrade feet (Tuttle, 1967). Despite the categorization as a terrestrial quadruped, the mountain gorilla utilizes arboreal substrates for a variety of activities, including feeding, nesting (Pitman, 1935; Donisthorpe, 1958, Schaller, 1963;Jones and Sabater Pi, 1971; Fossey, 1979; Tuttle and Watts, 1985), and playing (Schaller, 1963; Fossey, 1979; Tuttle and Watts, 1985). Earliest accounts of such activities were anecdotal but have since been confirmed by quantitative behavioral studies (Tuttle and Watts, 1985).Not surprisingly, the frequency with which gorillas engage in arboreal activities, including climbing, has been shown to vary inversely with body size (Schaller, 1963; Tuttle and Watts, 1985). This relationship was identified initially by Schaller (1963), who observed age- and sex-related differences in tree-climbing and arm-swinging, and documented quantitatively more recently by Tuttle and Watts (19851, who note that adults knuckle-walk more frequently than either juveniles or young adults, younger (and smaller) gorillas are more arboreal compared to the extremely terrestrial silverbacks, and adult females utilize arboreal substrates more frequently than adult males. Among the various skeletal components associated with locomotion, the scapula has been examined extensively, perhaps because of its evolution a s part of the shoulder complex from a weight-bearing organ in terrestrial quadrupeds to a n organ of manipulation and balance in modern humans (Schultz, 1930; Coolidge, 1933; Ashton et al., 1965; Oxnard, 1963,1967,1976; Preuschoft, 1973; Siege1 and Jones, 1975; Kimes et al., 1979; Doyle et al., 1980; Shea, 1986). Variation in scapula morphology among primates has been primarily linked to differences in frequency of arboreal and terrestrial behaviors (Coolidge, 1933; Roberts, 1974; Oxnard, 1963, 1967). Many analyses of scapula proportions have suggested that the scapula reflects a distribution of shape change, with extreme arboreality in the gibbons, siamangs, and orang-utans at one end of the hominoid distribution and extreme terrestriality in gorillas and humans at the other. Gibbons and siamangs, for example, are brachiators characterized by marked vertical elongation of the scapula, relatively narrow supraspinous and infraspinous fossae, and a vertebral border obliquely sloped relative to the scapula spine. These features have been related to their increased ability to circumduct, to make rapid reaching movements a s they move between overhead supports, and to meet the tensile demands placed on the shoulder joint during constant overhead arm-swinging (Roberts, 1974). By contrast, the gorilla, a terrestrial quadrupedal knucklewalker, bears scapulae with very broad supraspinous and infraspinous fossae, both of which are often further triangulated into smaller fossae presumably reflecting increased muscle mass (Roberts, 19741,and a vertebral border nearly perpendicular to the scapula spine, designed to withstand compressive forces along the vertebral border (Roberts, 1974). Chimpanzees, intermediate in behavior and morphology, are described a s having narrower scapulae with less well-developed supraspinous fossae than gorillas (Roberts, 1974); and humans have broad, well-developed infraspinous fossae but poorly developed supraspinous fossae, reflecting the fact that humans do not habitually elevate the upper limb above the level ofthe shoulder (Roberts, 1974). Thus, it has been argued specifically that primates with relatively vertically elongate, narrower scapulae and higher scapular indices, while not necessarily converging on the hylobatid pattern of brachiation, should at least be better adapted to, and incorporate with relatively greater frequency, armswinging, arm-hanging, climbing, and other arboreal behaviors (Ashton and Oxnard, 1964; Ashton et al., 1965; Roberts, 1974; Horn, 1975; Susman, 1980, 1984). Locomotor behavior often changes with SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA body size both between and within species. For example, in their comparative study of New World monkeys, Fleagle and Mittermeier (1980) demonstrated that increase in body size among species is accompanied by a corresponding increase in arboreal climbing and decrease in leaping. Crompton's (1983) study of two species of galago yielded similar results. Doran's (1992) study of chimpanzee locomotor behavior revealed that pygmy and common chimpanzees undergo changes in locomotor and positional patterns during ontogeny. Other studies have documented a within-species relationship between differences in positional behavior and body size. Gebo (19921, for example, found that female Alouatta paliatta utilize smaller diameter supports and climb more frequently than larger males. Among the apes, Sugardjito and van Hooff (1986) and Cant (1987) showed that sexual dimorphism in body size correlates with differences in positional behavior between male and female orangutans. Doran (1993) found that sexual dimorphism in body size correlates with differences in positional behavior between adult male and female chimpanzees. Despite evidence of a within-species correlation between body size and locomotor behavior, studies establishing a within-species relationship among body size, locomotor behavior, and skeletal morphology in primates are lacking. Shea (1986) demonstrated that scapula proportions differ between subadult and adult chimpanzees, and these morphological differences have been purported t o correlate with age-related changes in locomotor behavior (Doran, 1992). If changes in body size influence the relationship between form and function, then such effects should be most apparent in a species like Gorilla which exhibits a wide range in body size during the life cycle. Therefore, this study investigates the effects of ontogeny and sexual dimorphism on scapula morphology in one subspecies of Gorilla (G.g. beringei).Specifically, this work evaluates whether sexual differences in scapula morphology are the results of ontogenetic scaling, whether scapula form varies with age, and whether any sex- or age-related differences in scapula form correlate with differences in locomotor behavior. 433 MATERIALS AND METHODDS Samples The scapulae and associated skeletal and cranial material for G.g. beringei were obtained from the collections of the Musee Royal de l'Afi-ique Centrale (Tervuren, Belgium), Musee de la Vie (Louvain-La-Neuve, Belgium), the National Museum of Natural History (Washington, DC), and the Comparative Museum of Zoology (Cambridge, MA). Measurements were taken entirely from wild-caught specimens of known locality. For comparisons of age-related changes, data are grouped into three age categories. Age was estimated by use of a combined dental and cranial classification utilizing associated crania as follows: stage 1, all deciduous teeth fully erupted, plus any combination of the following: M' partially or fully erupted, M2 partially or fully erupted, and C and M3 erupting-juvenile (n = 9); stage 2, full permanent dentition, basilar suture open, moderate wear-subadult (n = 11); and stage 3, full permanent dentition, basilar suture closed, heavy wear-adult (n = 18). This aging scheme has been utilized in previous analyses (Korgman, 1931; Randall, 1943; Ashton and Zuckerman, 1950; Shea, 1981) and found to be reliable (Taylor, 1992). Estimation of age of individuals of known sex but lacking associated crania is by stepwise discriminant function analysis utilizing partial skeletal weight (computed from the combined weights of the skull, mandible, paired femora, humeri, scapulae, and pelvis, including the sacrum [Steudel, 1981]), humeral length, femoral length, iliac length, and femoral circumference. Percent of grouped cases classified correctly is 95.0% for males and 100% for females. Measurements A total of six linear measurements was obtained on each scapula (Schultz, 1930; Shea, 1986; Takahashi, 1990) (Fig. 1). These include 1) maximum scapula length, measured from the middle of the dorsal border of the glenoid fossa to the point on the vertebral margin where the medial end of the scapula spine directly meets the vertebral border (AD), 2) total breadth, measured as the maximum straight line distance from the superior 434 A.B. TAYLOR B 3. Infraspinous fossa lengthiscapula length x 100; 4. Scapula lengtldspine length x 100; 5 . Spine lengthisupenor border length X 100; and 6. Supraspinous fossa lengtldinfraspinous fossa length x 100. Methods of analysis Two-way analyses of variance (ANOVA) were carried out to test for significant sexand age-related differences in size and shape of the scapula, as represented by linear measures of size and indices of shape. Each analysis treated sex and age a s independent variables and a measure of scapula size or shape the dependent variable (Fig. 2). Tukey’s E as HSD post-hoc test was used to protect from Fig, 1. Drawing of an adult male Gorilla gorilla be- declaring pairs of means different when they ringei. Measurement definitions described in text. Re- could differ by chance, and, because counts drawn from Schultz (1930). of cases per cell were unequal, the harmonic mean was utilized by applying the TukeyKramer adjustment. Results of these and all border of the scapula to the inferior border of other analyses are considered statistically the scapula (B-E), 3) spine length, measured significant at the P < 0.05 level. All statistifrom the point on the vertebral margin cal analyses were carried out utilizing SYSwhere the medial end of the scapula spine TAT (1992). directly meets the vertebral border to the Bivariate ordinary least squares (OLS) most distant point on the acromion process and reduced major axis (RMA) regression (A-C), 4) supraspinous fossa length, mea- analyses were carried out on log-transsured from the tip of the superior border to formed data to describe and compare allothe point on the vertebral margin where the metric growth trajectories between sexes. medial end of the scapula spine and the ver- Analysis of covariance (ANCOVA) was used tebral border meet (A-B), 5) infraspinous to test for significant differences in patterns fossa length, measured from the tip of the of relative growth between OLS regressions inferior border of the scapula to the point on for the sexes. The method described by the vertebral margin where the medial end Clarke (1980) was used to test for significant of the scapula spine and the vertebral border RMA slope differences between sexes and meet (A-E), and 6) superior border length, Tsutakawa and Hewett’s (1977) “quick test” defined here as the distance between the su- used to test for significant positional differperior border of the scapula and the tip of ences in RMA regression lines. Because samthe acromion process (B-C). All linear mea- ple sizes are small and there are only six surements were recorded with a digital cali- bivariate comparisons, only differences idenpers accurate to 0.01 mm. Data were input tified by both line-fitting techniques were directly into a portable personal computer considered statistically significant at P < via a GagePort computer interface. 0.05. From linear measures, six shape indices were generated: RESULTS Scapula size Tests of statistical differences in scapula 1. Scapula lengthiscapula breadth X 100; 2. Supraspinous fossa lengthhcapula size by sex and age are all significant (Table 1). Post-hoc tests reveal significant differlength x 100; 435 SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA JUVENILE SUBADULT ADULT d ? Fig. 2. A two-by-three model depicting sex and age as the independent variables and scapula shape as the dependent variable. Adapted from Schultz (1930). ences between sexes for all scapula dimensions a t the subadult and adult stages. At the juvenile stage, males and females are characterized by scapulae of comparable sizes (Table 2). Of note is that juvenile females are (not significantly) larger than their male counterparts for all measures of scapula size. This pattern is reversed in subadults, where males are characterized by significantly larger scapulae than females of comparable age. Thereafter, males remain larger than females of comparable age for all measures of scapula size. As expected, sexual dimorphism in scapula size becomes more extreme as body size and age increase. Descriptive growth statistics for each linear measurement by age indicate t h a t growth from juvenile to adult proceeds at a fairly constant rate for all measures of scapula size, with juvenile males achieving approximately 45.0% and juvenile females approximately 60.0% of their full adult sizes (Table 3). In all cases, growth from subadult to adult is negligible, involving less than a 10% increase in size in all cases. Scapula shape by sex and age Tests of statistical differences in scapula shape by sex and age are significant only for age (Table 4). Specifically, shape ratios of scapula length vs. scapula breadth, infraspinous fossa length vs. scapula length, and scapula length vs. spine length vary significantly with age. In comparison to adults, post-hoc tests reveal that juveniles have significantly higher index values for scapula length vs. scapula breadth and significantly lower values for infraspinous fossa length vs. scapula length (Table 5). Juveniles have comparatively shorter scapula spines compared with either subadults or adults. Relative growth of the scapula Statistical tests of slope and position differences reveal that, for the most part, males and females are characterized by ontogenetic scaling. In other words, males and females share a common growth trajectory such that males continue to grow along a common growth vector after growth in fe- 436 A.B. TAYLOR TABLE 1. Means (mm), standard deviations (Sol, sample sizes, and results of tests for statistical differences in six Linear measures of scapula size for male and female Gorilla gorilla beringei' Variable Sex Juvenile Age Subadult Adult Significance by sex Male 81.7 (21.9) 3 91.0 (9.9) 6 169.5 (16.9) 7 131.2 (10.9) 4 178.2 (10.5) 12 133.1 (6.4) 6 110.6 (30.2) 118.9 (19.7) 237.0 (26.3) 180.9 (16.6) 253.7 (19.5) 190.9 (10.7) 62.5 (18.4) 63.1 (9.5) 129.0 (13.2) 97.4 (12.7) 136.3 (14.7) 102.4 (7.1) 60.1 (16.7) 64.2 (12.6) 129.3 (17.9) 99.9 (8.7) 137.9 (8.1) 102.2 (4.7) 97.0 (26.6) 108.3 (14.0) 221.3 (24.5) 171.8 (16.3) 229.8 (18.6) 176.8 (11.0) 54.2 (11.2) 66.4 (7.3) 131.2 (18.9) 102.4 (14.0) 130.9 (13.7) 105.5 (9.2) Scapula length Female ** Scapula breadth Male Female Supraspinous fossa length Male Female Infraspinous fossa length Male Female Spine length Male Female ** Superior border length Male Female * 'Results are for mean differences between sexes at each age. ' P i 0.05. **P < 0.001. TABLE 2. Results of post-hoc tests for statistical differences in six linear measures of scapula size between .sexes for each age group Male vs. female iuveniles' Variable Male vs. female subadults Male vs. female adults NS NS NS NS NS NS Length Breadth Supraspinous fossa length Infraspinous fossa length Spine length Sunerior border leneth INS, nonsignificant. * P i 0.05. TABLE 3. Descriptiue growth statistics by age and sex for six linear measures of the scapula' Variable Male Scapular length Scapular breadth Supraspinous fossa length Infraspinous fossa length Spine length Sunerior border leneth 45.8 43.7 45.9 43.6 42.2 41.4 Juvenile Female 68.4 62.3 61.6 62.8 61.3 62.9 'Values represent percentage of growth as computed with adult as base. Male 95.1 93.4 94.6 93.8 96.3 100.7 Subadult Female 98.6 94.8 95.1 97.7 97.2 97.1 Adult Male Female 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA 437 TABLE 4. Means, standard deviations (SD),sample sizes, and results of tests of statistical differences in scapula shape by age and sex in Gorilla gorilla beringei’ Variable* Sex Juvenile Age Subadult Male 74.2 (6.10) 3 77.4 (7.9) 6 71.7 (5.0) 7 72.7 (4.6) 4 70.4 (2.8) 12 69.8 (3.5) 6 76.2 (5.5) 69.3 (6.6) 76.1 (2.2) 74.3 (8.2) 76.3 (5.0) 77.0 (5.5) 73.7 (9.7) 70.2 (7.3) 76.4 (8.2) 76.2 (2.5) 77.5 (3.9) 76.8 (1.8) 84.3 (2.0) 84.5 (8.3) 76.7 (2.6) 76.5 (4.8) 77.8 (3.6) 75.3 (1.5) 56.5 (3.8) 61.7 (5.7) 59.2 (4.4) 59.6 (5.7) Significance by age Significance by sex * NS NS NS * NS ** NS 57.0 (4.0) 59.7 (3.7) NS NS 98.7 (8.5) 100.4 (7.9) NS NS Adult SCWSCB x 100 Female SFWSCL x 100 Male Female IFWSCL x 100 Male Female SCWSPL x 100 Male Female SBWSPL x 100 Male Female SFMFL x 100 Male 103.9 1(00.5 (9.6) 97.5 (9.3) (6.8) Female 99.1 (7.8) ‘Tests for significant differences among age groups are with sexes combined. ‘SCL, scapula length; SCB, scapula breadth; SFL, supraspinous fossa length; IFL, infraspinous fossa length; SPL, spine length; SBL, superior border length. * P C 0.05.**P < 0.01. NS, nonsignificant. TABLE 5. Results of post-hoc tests for statistical differences in scapula shape by age’ Juvenile vs. subadult Scapula length vs. scapula breadth Infraspinous fossa length vs. scapula length Scapula length vs. spine length NS NS * Juvenile vs. adult * Subadult vs. adult * NS NS * NS ’ NS, nonsignificant. * P C 0.05. males has ceased (Tables 6, 7, Fig. 3A-F). Visual examination of bivariate plots reveals that patterns of growth for shape comparisons between males and females are fairly linear (Fig. 3A-F). For all bivariate comparisons save one, RMA regression analyses yield higher slopes and lower y-intercepts t h a n OLS regression analyses, the exception being the slope and intercept comparison for males of spine length vs. superior border length. The trend is for females to scatter below males (Fig. 3A-F). In all cases, correlation coefficients are sufficiently high so that RMA slopes fall within the 95% confidence intervals of the OLS slopes. Results of Clarke’s test for slope differences on RMA regressions did not always concur with results of ANCOVA on OLS regressions (Table 7). Specifically, ANCOVA reveals significant sex differences in slopes of bivariate plots of scapula breadth against scapula length and scapula length against 438 A.B. TAYLOR TABLE 6. OLS and RMA regression statistics RMA OLS Male Female Scapula length vs. scapula breadth 0.919 Slope 0.042 y-intercept (0.847-1.051) 95% c1 R 0.98 Scapula length vs. supraspinous fossa length Slope 0.944 y-intercept 0.235 (0.867-1.021) 95% CI R 0.98 Scapula length vs. infraspinous fossa length Slope 0.891 v-interceut 0.344 (0.771-1.011) i 5 % CI R .. R 0.888 0.152 (0.840-0.936) 0.99 Supine length vs. superior border length Slope 0.948 y-intercept 0.348 9 5 8 Cl (0.853-1.043) R 0.98 Supraspinous fossa length vs. infraspinous fossa length Slope 0.927 y-intercept 0.150 95% R c1 (0.811-1.043) 0.96 Female 0.929 0.018 (0.849-1.016) 0.833 0.228 (0.729-0.952) 0.772 0.596 (0.611-0.913) 0.94 1.000 0.120 (0.914-1.094) 0.833 0.462 (0.691-1.004) 0.767 0.582 (0.680-0.854) 0.98 0.929 0.269 (0.796-1.083) 0.833 0.454 (0.747-0.930) 0.802 0.330 (0.658-0.946) 0.95 0.929 0.055 (0.871-0.990) 0.909 0.097 (0.766-1.079) 0.991 0.236 (0.819-1.163) 0.95 0.933 0.378 (0.853-1.021) 1.000 0.220 (0.843-1.186) 0.925 0.139 (0.774-1.076) 0.96 0.929 0.149 (0.818-1.054) 1.000 -0.010 (0.858-1.166) 0.717 0.333 (0.670-0.904) 0.97 0.96 . .~ Scapula length vs. spine length Slope y-intercept 95% CI Male TABLE 7. Results of OLS and RMA tests for slope and y-intercept differences between sexes' RMA OLS Scapula length vs. scapula breadth Scapula length vs. supraspinous fossa length Scapula length vs. infraspinous fossa length Scapula length vs. spine length Spine length vs. superior border length Supraspinous vs. infraspinous fossa length SloDe v-interceDt SloDe v-interceDt * NS NS NS NS NS NS NS NS NS NS NS NS * NS NS NS NS * NS NS * NS NS NS, nonsignificant * P < 0.05. supraspinous fossa length, whereas Clarke's test yields nonsignificant differences in slopes for all six bivariate regression analyses. ANCOVA also reveals a significant difference between the sexes for the y-intercept for the bivariate comparison of superior border length against spine length. Again, Tsutakawa and Hewett's (1977) significance tests for position differences do not concur with ANCOVA on this significant result but rather yield a significant position difference for the comparison of scapula length vs. spine length. Since significant slope and intercept differences could not be substanti- ated by both methods, results are interpreted as reflecting a pattern of similarity in scapula shape between sexes. DISCUSSION As expected, data presented in this investigation reveal significant differences in scapula size between males and females during ontogeny. Previous studies of common chimpanzees (Gavan, 1953; Smith et al., 1975; Shea, 1981) and gorillas (Randall, 1943; Shea, 1981; Leigh, 1992) reveal that females tend to be larger than males at early 439 SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA 2.6 2.60 I I I A 2.4 2.35 5m 2 m :2.2 -aa 0 I =n rn 2.10 a V u) 3 m m 2.0 -1 -I 1.85 1.8 I 1.60 1.6 I I 1.9 2.2 2.5 1.6 1.50 2.8 I I 1.94 2.16 I 2.38 2.60 Log Supraspinous Fossa Length Log Scapula Breadth 2.6 I 1.72 2.6 1 1 I D 2.4 - 5m 5 m f -anm s 3 2.2 - -am n s 2.0 - m 2.2 - 2.0 - 0 -1 -1 1.8 - 1.8 1.6 1.50 1.72 1.94 2.16 2.38 - ' I 1.6 1.70 2.60 I I , 5 2.58 2.80 t I 1 2.36 m rn -1 w 9) I z 2.16 U. 2.12 m m 0 L 1.94 1.88 E 3 u) n m 3 I 2.36 F 5 2.38 -f , 2.60 I E 2 2.14 Log Splne Length Log lntrasplnous Fossa Length 2.00 I I 1.92 u) k 1.72 1.64 -1 1.so 1.70 , 1.92 I 2.14 I I I 2.36 2.58 2.80 Log Spine Length 1.40 1.5 I I I 8 1.7 1.9 2.1 2.3 Log lntrasplnous Fossa Length Fig. 3. A-F: Bivariate plots of linear regressions of scapula dimensions for Gorilla gorilla beringei and 95% confidence ellipses. All variables are in base 10 logarithms. M, males: F, females. Associated regression statistics presented in Tables 6 and 7. 2.5 440 A.B. TAYLOR stages of growth but that males grow faster and for longer duration than females. Patterns of scapula growth identified in this study appear consistent with previous findings for the African apes. Significant differences between subadults in all dimensions of the scapula suggest that male G.g. beringei achieve larger adult body size by growing at a faster rate than females (Table 2) (Gould, 1977; Shea, 1983, 1988; Ravosa et al., 1993; Ravosa and Ross, 1994). This pattern is evident, for example, in the increase in scapula length. The average subadult male scapula is 169.5 mm in length, which is approximately 95% of its average adult length of 178.2 mm. The average subadult female scapula is 131.2 mm in length, approximately 99% of its average adult length of 133.1 mm. Therefore, males grow a t a faster rate than females because they undergo a greater increase in length compared to juvenile females and as adults they have longer scapulae than females. The marked increase in differences in scapula size from subadult to adult suggests that males also attain larger adult body size by growing for longer duration; that is, males continue to grow after growth in females is complete. Although not statistically significant, juvenile females achieve a greater percentage of their adult scapula size compared to juvenile males. Interestingly, similar data for lowland gorillas for which data on infants and larger sample sizes were available reveal nonsignificant differences between infant males and females and that males are consistently larger than females throughout growth (Taylor, 1992, in preparation). One explanation for these subspecies differences, and for the discrepancy between previous findings of growth for the African apes and results for lowland gorillas, is that the aging scheme utilized in this investigation for identifying juveniles encompasses too many distinct growth stages. As such, the early growth spurt typical of females would be obscured. However, since the aging criteria were the same for both subspecies, this interpretation seems unsatisfactory. Rather, such findings are either suggestive of true differences in patterns of growth between subspecies, or the results of small, biased sample sizes for juvenile mountain gorillas. Larger sample sizes and data on infants for the mountain gorillas will undoubtedly clarify this issue. Age-related differences in scapula form One of the original premises of this investigation was that changes in body size ought to affect the relationship between form and function, meaning that in order for animals of differing size to accomplishthe same kinds of tasks, changes in form must accompany an increase or decrease in body size. It was further argued that such changes should be more dramatic in species characterized by extreme ranges in body size-not simply between sexually dimorphic individuals, but among individuals of varying size within a single ontogeny. Data from this investigation reveal three significant age-related shape differences involving ratios of scapula length vs. scapula breadth, infraspinous fossa length vs. scapula length, and scapula length vs. spine length (Table 4). All three shape ratios differ significantly between juveniles and adults, and the shape ratio of scapula length vs. spine length differs between juveniles and subadults as well. Larger sample sizes and data on infants would likely increase the number of significant age-related differences, as was the case for similar data on lowland gorillas (Taylor, 1992, in preparation). Mean scapula length for juveniles is approximately 82.0 mm and mean scapula breadth approximately 111.0 mm (Table 1). These lengths are essentially doubled in subadults, where mean length is 170.0 mm (an increase of about 107.0%)and mean breadth 237.0 mm (an increase of approximately 114.0%).These two values have more than doubled, however, since increase in scapula length and breadth by exactly 100% would result in values of 164.0 mm and 222.0 mm, respectively. Mean length exceeds an increase of 100% by 6.0 mm, mean breadth by 15.0 mm. In other words, as size increases scapula breadth not only remains longer than scapula length, but its rate of increase is greater relative to that of scapula length, resulting in subadults with absolutely and relatively broader scapulae compared to juveniles. It is important to note that a rela- SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA tively broader scapula is more vertically elongated (that is, longer and narrower perpendicular to the scapula spine). The longer infraspinous fossa length relative to scapula length in juveniles compared to adults reflects this vertical elongation (Table 3). At larger body sizes, scapula breadth continues to grow a t a faster rate than scapula length, resulting in the well-established pattern for gorillas of scapulae with markedly broad fossae, in particular with relatively broader supraspinous fossae compared to any other primate (Roberts, 1974). Rate of increase along both dimensions slows considerably a s gorillas approach full adult size. Adult male scapulae are only about 5% longer and 7% broader than subadults. The pattern is the same for females, except percentage of growth from subadult to adult is less than for males, confirming that females grow for shorter duration (Table 3). Variation in length vs. breadth index, referred to by Schultz (1930) as the “scapula index,” has been linked to various functional differences in behavior among primates. I t has been well established that high index values for the ratio of scapula length vs. scapula breadth are characteristic of proportionately narrow scapulae, as are low indices for infraspinous fossa length vs. scapula length. Ashton and Oxnard (1963) have suggested that narrow and elongated scapulae, such as are characteristic of the gibbon and siamang, provide increased mechanical advantage to scapula stabilizers and rotators, such as the serratus anterior and trapezius muscles. They argue the comparatively high index of the gibbons affords these animals greater range of motion and increased joint stability, mechanical advantages necessary for withstanding tensile forces involved in grasping overhead supports during mid-air propulsion (Ashton and Oxnard, 1963). Roberts (1974) has echoed this claim, suggesting that the gibbon’s narrow scapula is designed to provide increased circumduction and joint stabilization to the shoulder. More recently, it has been suggested that the long, narrow scapula of the chimpanzee is related to a n increase in ability to abduct and upwardly rotate the humerus during unimanual arm-hanging, thereby achieving a relatively close approximation of the gleno- 44 1 humeral joint to the spinal column and reducing the mechanical stresses associated with arm-hanging behavior (Hunt, 1991). These hypotheses focus on the mechanical advantages afforded the serratus anterior and trapezius muscles, and the concomitant reduction in mechanical stresses, and suggest that arm-hanging and brachiation are kinematically similar behaviors requiring similar bony and muscular adaptations. Values €or the scapula index in mountain gorillas range from about 77.0 for juveniles to 70.0 for adults (Table 4). Of interest are data on lowland gorillas for which individuals of dental stage 1(infants)were available (Taylor, in preparation), revealing a higher scapula index for infants compared to juveniles (82.0 for infants as compared to 74.0 for juveniles) and increasing the range for this index by about 4%. A reasonable inference is that data on G.g. beringei infants would similarly increase the range for mountain gorillas and result in a higher frequency of significant age-related differences. One could predict that a high scapula index would be associated with behaviors such as suspensory locomotion, arm-hanging, arm-swinging, and, perhaps, vertical climbing, whereas a relatively low index would be associated with more terrestrial quadrupedal behaviors. Data from this study indicate that juveniles are characterized by relatively higher scapula indices than either subadults or adults, suggesting that younger gorillas should be incorporating a greater component of arm-hanging behaviors than at any other stage of growth-a prediction based not only on the assumption that small body size facilitates activities which are prohibitive a t larger body sizes but that scapula form in younger individuals is better suited to such behaviors. In fact, these predictions are borne out by behavioral data which indicate that frequency of arboreal activity is higher in younger, smaller individuals than in older, larger ones (Schaller, 1963; Goodall, 1977; Tuttle and Watts, 1985). In this study, data reveal that subadults and adults have significantly longer scapula spines compared to scapula lengths than juveniles (Tables 1,5). Schultz (1930) has suggested the length vs. spine length index reflects the relative extent to which the 442 A.B. TAYLOR acromion process extends beyond the level of the glenoid fossa, postulating that a relatively longer spine and acromion process provide increased protection of the shoulder joint. Roberts (1974) hypothesized that the acromion functions in a bridge-like manner to provide structural support and increased leverage for the deltoid muscle. More recently, i t has been suggested (Takahashi, 1990) that a well-developed scapula spine is logically a function of increased muscle attachment of the trapezius and deltoid muscles, thus a t least partially echoing Roberts’s earlier claim. Takahashi’s results led her to conclude that a well-developed spine in Hylobates functions to increase the mechanical advantage of these muscles in elevating the humerus during arm-swinging, although she admits that such muscles are equally important in vertical climbing as well. Therefore, the presence of a well-developed scapula spine in the mountain gorilla may serve a dual advantage. First, a welldeveloped scapula spine provides a n increased mechanical advantage in elevating the humerus during arm-hanging activities, especially in younger individuals with lessdeveloped associated muscles, articulations, and ligaments to reinforce the glenohumeral joint. Second, the mechanical advantage of the trapezius and deltoid muscles is enhanced, particularly in larger, heavier individuals who not only vertically climb, but who frequently use their upper limbs in a n abducted position during feeding (Roberts, 1974; Tuttle and Watts, 1985). Tuttle and Watts (1985) found a n inverse relationship between forelimb length and frequency of humeral protraction above the shoulder joint during feeding bouts in the mountain gorilla. These data support the hypothesis that increased leverage of muscles associated with humeral abduction is particularly advantageous in larger individuals. Larson and Stern (1986) have demonstrated that electrical potentials of the supraspinatus and deltoid muscles in chimpanzees are highest during the initial phases of humeral elevation, during which these muscles are active in resisting the tendency of the deltoid muscle to displace the humerus superiorly. Recent studies in chimpanzee positional behavior (Hunt, 1991,1992) indicate that chimpanzees require a fully abducted humerus during arm-hanging and large muscles for arm-raising, such as teres minor, cranial trapezius, anterior deltoid, and clavicular portion of pectoralis major (Hunt, 1991). These data also strengthen the suggestion that a relatively long scapula spine is mechanically advantageous in humeral elevation during climbing and hanging activities. Sex differences in scapula form The pervasive pattern for mountain gorillas is one of similarity in scapula form between sexes. Comparisons of male and female regressions suggest that differences in scapula proportions between adults result from extension of common patterns of relative growth (Tables 4, 7; Fig. 3A-F). With the exception of the two shape indices involving the length of the spine (Fig. 3D-E), the trend is for males and females to overlap at smaller sizes and diverge through time, with females shifting below males and males continuing to grow to larger terminal sizes. The effects of sexual dimorphism on locomotor (Fleagle and Mittermeier, 1980; Doran, 1993),positional (Cant, 1987; Doran, 1993), and feeding behaviors (Wrangham and Smuts, 1980; Galdikas and Teleki, 1981; Cant, 1987) have been well documented. By contrast, few studies have correlated behavioral differences between sexes with genderspecific morphologies. This study suggests that, despite differences in locomotor behavior between sexes, there is little evidence to suggest that males and females have evolved unique morphological adaptations to different ecological niches. Results of this investigation also raise the issue of whether frequency or kinematics (Prost, 1965, 1967) is the relevant variable in correlating morphology with behavior. This issue is particularly germane to studies attempting to identify gender-specific morphological patterns, since the “average biomechanical situation” (Oxnard, 1979) will likely be more similar between sexes than between species, potentially obscuring subtle differences. Although tempting to attribute the absence of significant proportion differences between sexes to small sample sizes, a similar analysis of sexual differences for lowland gorillas for which SCAPULA MORPHOLOGY IN THE MOUNTAlN GORILLA sample sizes were notably larger yielded results similar to those found for the mountain gorilla (Taylor, 1992, in preparation). SUMMARY AND CONCLUSIONS The purpose of this investigation was to assess the effects of ontogeny and sexual dimorphism on scapula form in the mountain gorilla and to correlate any differences in scapula morphology with variability in locomotor behavior. Data reveal significant agerelated variability in scapula form. However, sexual differences in shape of the scapula can be explained by ontogenetic scaling of scapula proportions. Ontogenetic differences in scapula form in gorillas are suggestive of locomotor differences relating to frequency of arboreal vs. terrestrial behaviors. Thus, younger and more actively arboreal gorillas appear to fit a morphological pattern which more closely approximates the hylobatid pattern of vertically elongated and mediolaterally compressed scapulae. Previous comparisons of scapula form in chimpanzees have revealed significant proportion differences between adult common and pygmy chimpanzees which reflect form differences between subadult and adult common chimpanzees (Shea, 1986). Given the proportion differences among mountain gorillas of varying agelsize, it would be worthwhile to determine if differences in scapula form between subspecies reflect the specific proportion differences documented among age groups. Such data would enable us to distinguish between genetic-based differences in scapula form and those related to remodelling during growth. Behavioral data on the lowland gorilla are emerging from the field (Remis, 1994,1995). In addition, recent genetic data strongly support a greater degree of separation between the two subspecies of gorilla than had been previously accepted (Ruvolo e t al., 1991; Ruvolo, 1994). These data suggest the need for a more comprehensive investigation of variability in locomotor morphology, including the scapula, both within and between species. Such data would enhance our understanding of patterns of variability and lead to a fuller appreciation of the functional 443 implications of morphological variability among the African apes. ACKNOWLEDGMENTS I thank M.I. Siegel, B.T. Shea, and M.D. Rose for helpful comments and suggestions. This manuscript benefited greatly from the comments and criticisms of two anonymous reviewers. I am grateful for the support and assistance of Dr. W. Van Neer, Dr. R.W. Thorington, Ms. M. Rutzmoser, and the staffs of the Musee Royale de l’Afrique Centrale, the National Museum of Natural History, and the Museum of Comparative Zoology. I am particularly indebted to Bernard and MariePaul Latteur (MusBe de la Vie) for their gracious hospitality during my stay in Belgium. This research was supported by the National Science Foundation (BNS 90-16522), Wenner-Gren Foundation (Gr. 53231, Sigma Xi Grants-in-Aid of Research, and a Mellon Predoctoral Fellowship award. LITERATURE CITED Ashton EH, and Oxnard, CE (1963) The musculature of the primate shoulder. Trans. Zool. SOC.Lond. 29:553650. Ashton EH, and Oxnard CE (1964) Functional adaptations in the primate shoulder girdle. Proc. 2001. SOC. Lond. 142:49-66. Ashton EH, and Zuckerman S (1950) Some quantitative dental characteristics of the chimpanzee, gorilla, and orang-utan. Philos. Trans. R. SOC. Lond. [Biol.] 234: 471-484. Ashton EH, Oxnard CE, and Spence TF (1965) Scapular shape and primate classification. Proc. Zool. SOC. Lond. 145:125-142. Cant J G (1987)Effects of sexual dimorphism in body size on feeding postural behavior of Sumatran orangutans (Pongo pygmaeus). Am. J. Phys. Anthropol. 74:143148. Clarke MRB (1980)The reduced major axis of a bivariate sample. Biometrika 67:441-446. Coolidge H J (1933) Pan paniscus: Pygmy chimpanzee from south of the Congo River. Am. J. Phys. Anthropol. 18:l-57. Crompton RH (1983) Age differences in locomotion of two subtropical Galaginae. Primates 24:241-259. Donisthorpe J H (1958) A pilot study of the mountain gorilla (Gorilla gorilla beringei) in South West Uganda, February to September. S. Afr. J. Sci. 54:195217. Doran DM (1992) The ontogeny of chimpanzees and pygmy chimpanzee locomotor behavior: A case study of paedomorphism and its behavioral correlates. J. Hum. Evol. 23:139-157. Doran DM (1993) Sex differences in adult chimpanzee positional behavior: The influence of body size on loco- 444 A.B. ‘I ‘AYLOR motion and posture. Am. J . Phys. Anthropol. 91:99115. Doyle WJ, Siegel MI, and Kimes KR (1980) Scapular correlates of muscle morphology in Macaca mulatta. Acta. Anat. 106:493-501. Fleagle, JG (1988) Primate Adaptation and Evolution. San Diego: Academic Press, Inc. Fleagle JG, and Mittermeier RA (1980) Locomotor behavior, body size, and comparative ecology of seven Surinam monkeys. Am. J . Phys. Anthropol. 52:301314. Fossey D (1979) Development of the mountain gorilla (Gorilla gorilla beringei): The first thirty-six months. In DA Hamburg and ER McCown (eds.): The Great Apes. Menlo Park: BenjamidCummings Publishing co. Galdikas BMF, and Teleki G (1981)Variations in subsistence activities of female and male pongids: New perspectives on the origins of hominid labor division. Curr. Anthropol. 22:241-256. Gavan JA (1953) Growth and development of the chimpanzee: A longitudinal and comparative study. Hum. Biol. 25:99-143. Gebo DL (1992)Plantigrady and foot adaptation in African Apes: Implications for hominoid origins. Am. J. Phys. Anthropol. 89:29-58. Goodall A (1977) Feeding and ranging behaviour of a mountain gorilla group (Gorilla gorilla beringei) in the Tshibinda-Kahuzi Region. In TH Clutton-Brock (ed.): Primate Ecology: Studies of Feeding and Ranging Behaviour in Lemurs, Monkeys, and Apes. London: Academic Press, pp. 449479. Gould SJ (1977) Ontogeny and Phylogeny. Cambridge: Harvard University Press. Horn AD (1975) Adaptations of the pygmy chimpanzee (Pan paniscus) to the forests of the Zaire Basin. Am. J. Phys. Anthropol. 42:307. Hunt K (1991) Mechanical implications of chimpanzee positional behavior. Am. J . Phys. Anthropol. 86:521536. Hunt K (1992) Positional behavior ofPan troglodytes in the Mahale Mountains and Gombe Stream National Parks. Am. J. Phys. Anthropol. 87:83-105. Inouye, SE (1992) Ontogeny and allometry of African ape manual rays. J . Hum. Evol. 23:107-138. Jones C, and Sabater Pi J (1971) Comparative ecology of Gorillagorilla (Savage and Wyman) and Pan troglodytes (Blumenbach)in Rio Muni, West Africa. Bibliot. Primatol. 13:l-96. Jungers WJ, and Cole M S (1992) Relative growth and shape ofthe locomotor skeleton in lesser apes. J. Hum. Evo~.23:93-105. Jungers WJ, and Susman RL (1984)Body size and skeletal allometry in African Apes. In RH Susman (ed.): The Pygmy Chimpanzee. Plenum Press: New York, pp. 131-177. Kimes K, Doyle WJ, and Siegel MI (1979) Scapular correlates of muscle morphology in Papio cynocephalus. Acta Anat. 104:41&420. Krogman WM (1931) Studies in growth changes in the skull and face in anthropoids 111. Growth changes in the skull and face of the gorilla. Am. J . Anat. 47:89-115. Larson SG, and Stern J T Jr (1986) EMG of scapulohum- era1 muscles in the chimpanzee during reaching and “arboreal” locomotion. Am. J. Phys. Anthropol. 176: 171-190. Leigh S (1992) Patterns of variation in the ontogeny of primate body size dimorphism. J. Human Evol. 23:27-50. Oxnard CE (1963) Locomotor adaptation in the primate forelimb. Symp. Zool. 27,324-345. Oxnard CE (1967)The functional morphology of the primate shoulder as revealed by comparative anatomical, osteometric, and discriminant function techniques. Am. J. Phys. Anthropol. 26:219-240. Oxnard CE (1976) Primate quadrupedalism: Some subtle structural correlates. Yrbk. Phys. Anthropol. 20: 438-554. Oxnard CE (1979)Some methodological factors in studying the morphological-behavioral interface. In ME Morbeck, H Preuschoft, and N Gomberg (eds.): Environment, Behavior, and Morphology: Dynamic Interactions in Primates. New York: Gustav Fischer, pp. 183-207. Pitman CRS (1935) The gorillas of the Kayonsa region, Westerin Kigezi, S.W. Uganda. Proc. Zool. SOC. Lond. 105:477494. Preuschoft H (1973) Functional anatomy of the upper extremity: In GH Bourne (ed.): The Chimpanzee, Vol. 6. Basel: Karger. Prost J (1965) A definitional system for the classification of primate locomotion. Am. Anthropol. 67:1198-1214. Prost J (1967) Bipedalism of man and gibbon compared using estimates ofjoint motion. Am. J. Phys. Anthropol. 26t135-148. Randall FE (1943)The skeletal and dental development and variability of the gorilla. Hum. Biol. 15:236-254. Ravosa MJ, and Ross CF (1994) Craniodental allometry and heterochrony in two howler monkeys: Alouatta seniculus and A. palliata. Am. J . Primatol. 33:277299. Ravosa MJ, Meyers DM, and Glander KE (1993) Relative growth of the limbs and trunk in sifakas: Heterochronic, ecological, and functional considerations. Am. J . Phys. Anthropol. 92:499-520. Remis MJ (1994) The effects of body size and social context on the positional behavior of lowland gorillas (Gorilla gorilla gorilla) in the Central African Republic. Am. J. Phys. Anthropol. (Suppl.) 18:167-168. Remis MJ (1995) Effects of body size and social context on the arboreal activities of lowland gorillas in the Central African Republic. Am. J . Phys. Anthropol. 97t413-433. Roberts D (1974) Structure and function of the primate scapula. In FA Jenkins Jr (ed.): Primate Locomotion. New York: Academic Press, pp. 171-200. Ruvolo ME (1994)Molecular evolutionary processes and conflicting gene trees: The hominoid case. Am. J. Phys. Anthropol. 94:89-113. Ruvolo ME, Disotell TR, Allard MW, Brown WM, and Honeycutt RL (1991) Resolution of the African hominoid trichotomy by use of a mitochondria1 gene sequence. Proc. Natl. Acad. Sci. U.S.A. 88:1570-1574. Schaller GB (1963) The Mountain Gorilla: Ecology and Behavior. University of Chicago Press, Chicago. Schaller GB (1965)The behavior ofthe mountain gorilla. In I DeVore (ed.): Primate behavior: Field Studies of SCAPULA MORPHOLOGY IN THE MOUNTAIN GORILLA Monkeys and Apes. New York Holt, Rinehart and Winston, pp. 324-367. Schultz AH (1930) The skeleton of the trunk and limbs of higher primates. Hum. Biol. 2:303438. Shea BT (1981)Relative growth of the trunk and limbs of the African Apes. Am. J. Phys. Anthropol. 56:179-202. Shea BT (1983)Allometry and heterochrony in the African Apes. Am. J . Phys. Anthropol. 62:275-289. Shea BT (1984) An allometric perspective on the morphological and evolutionary relationships between pygmy (Pan paniscus) and common (Pan troglodytes) chimpanzees. In RL Susman (ed.): The Pygmy Chimpanzee: Evolutionary Biology and Behavior. New York: Plenum Press, pp. 237-266. Shea BT (1986) Scapula form and locomotion in chimpanzee evolution.Am. J . Phys. Anthropol. 70:475-488. Shea BT (1988) Heterochrony in primates. In ML McKinney (ed.): Heterochrony in Evolution. New York Plenum Press, pp. 237-266. Shea BT (1992) Ontogenetic scaling of skeletal proportions in the talapoin monkey. J. Hum. Evol. 23:283307. Siege1 MI, and Jones CL (1975) The skeletal correlates of behavioral modification in the laboratory mouse (Mus rnusculus). Am. J . Phys. Anthropol. 42:141-144. Smith AFt, Butler TM, and Pace N (1975) Weight growth of colony-reared chimpanzees. Folia Primatol. 24: 29-59. Steudel K (1981) Body size estimators in primate skeletal material. Int. J. Primatol. 2:81-90. Sugardjito J , and van Hooff J (1986)Sex-age class differences in positional behavior of the Sumatran orangutan (Pongo pygmaeus abelii) in the Gunung Leuser National Park, Indonesia. Folia Primatol. 47:14-25. 445 Susman RL (1980) Acrobatic pygmy chimpanzees. Nat. Hist. 89t32-39. Susman RL (1984)The locomotor behavior ofPanpaniscus in the Lomako forest. In RL Susman (ed.): The Pygmy Chimpanzee. New York: Plenum Press, pp. 369-394. SYSTAT(1992)SYSTAT for Windows: Statistics, Version 5. Evanston, IL: SYSTAT, Inc., 1992. Takahashi LK (1990) Morphological basis of arm-swinging: Multivariate analyses of the forelimbs of Hylobates and Ateles. Folia Primatol. 54:70-85. Taylor AB (1992) A Morphometric Study of the Scapula in Gorilla (Gorilla gorilla gorilla and Gorilla gorilla beringei). Ph.D. dissertation, University of Pittsburgh. Taylor AB (in preparation) Comparative morphology of the scapula in lowland and mountain gorillas and correlates of locomotor behavior. Tsutakawa RK, and Hewett J E (1977) Quick test for comparing two populations with bivariate data. Biometrics 33:215-219. Tuttle RH (1967) Knuckle-walking and the evolution of hominoid hands. Am. J . Phys. Anthropol. 26:171-206. Tuttle RH, and Watts DP (1985)The positional behavior and adaptive complexes of Pan Gorilla. In S Kondo (ed.): Primate Morphophysiology, Locomotor Analyses and Human Bipedalism. Tokyo: University of Tokyo Press, pp. 261-288. Wrangham RW, and Smuts BB (1980) Sex differences in the behavioural ecology of chimpanzees in the Gombe National Park, Tanzania. J. Reprod. Fertil. Suppl. 28:13-3 1. Yerkes RM, and Yerkes AW (1929)The Great Apes. Yale University Press, New Haven.