SEARCH

SEARCH BY CITATION

Keywords:

  • knee;
  • Homo;
  • Pan;
  • Gorilla;
  • biomechanics

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

Although the hominid knee has been heavily scrutinized, shape variation of the medial tibial condyle has yet to be described. Humans, chimpanzees, and gorillas differ in the shape of their medial femoral condyles and in their capacity for external and internal rotation of the tibia relative to the femur. I hypothesize that these differences should be reflected in the shape of the medial tibial condyle of these hominids. Here I use geometric morphometric techniques to uncover shape differences between the medial tibial condyles of humans, chimpanzees, and gorillas. Humans are distinguished from the other two species by having a much more oval-shaped medial tibial condyle, while those of chimpanzees and gorillas are more triangular in outline. Gorillas (especially males) are distinguished by having more concavely-curved condyles (mediolateral direction), which is interpreted as an effect of heavy loading through the medial compartment of the knee in conjunction with differences in the degree of arboreality. Anat Rec, 296:1518–1525, 2013. © 2013 Wiley Periodicals, Inc.

Animal movement continues to be of keen interest to biologists and paleontologists alike because locomotion provides access to the key resources of food, water, safety, and potential mates. The articular surfaces of long bones are particularly useful for understanding locomotor adaptations because their size and shape are related to joint function, meeting the requirements for articular strength, mobility, and stability during normal locomotion (Currey, 1984; Swartz, 1989; Godfrey et al., 1991; Hamrick, 1996). Strength is the magnitude and frequency of loading a joint can withstand without suffering damage and is particularly important because irreparable damage can impede joint function and hinder locomotion (Hamrick, 1996). Joint mobility is the potential range of motion of a joint, and stability is the ability to withstand motions outside normal kinematics or that disrupt joint integrity (Hamrick, 1996).

The hominid knee is especially well-studied because of the information the joint provides for identifying bipeds in the fossil record and for understanding the evolution of hominin bipedalism (see Ward, 2002). Features of both the distal femur and proximal tibia have been used to distinguish bipedal humans from quadrupedal non-human apes (Thompson, 1889; Preuschoft, 1971; Tardieu 1981, 1983), fossil hominins from other extinct primates, and for taxonomic or locomotor distinctions among fossil hominins (Stern and Susman, 1983; Senut and Tardieu, 1985; Zihlman, 1985; Tardieu, 1986, 1999; McHenry and Berger, 1998). Compared to chimpanzees and gorillas, distinctive features of the human distal femur include a deep patellar groove with a projecting lateral lip (Le Gros Clark, 1947), distally flattened femoral condyles (Heiple and Lovejoy, 1971), and a high bicondylar angle (Stern and Susman, 1983). A higher bicondylar angle moves the knee closer to the midline of the body, making balancing on a single support foot easier (Preuschoft, 1971). The deep patellar groove prevents patellar subluxation/dislocation that can result from the contraction of the quadriceps muscle in combination with a valgus knee (Lovejoy, 2007). The distally flattened femoral condyles increase the contact area between tibial and femoral condyles at knee angles near full extension (Kettlekamp and Jacobs, 1972; Maquet et al., 1975; Maquet, 1976). All of these features are accommodations to habitual bipedal locomotion on a relatively extended lower limb.

Differences in knee function (and by extension locomotor behavior) are also reflected in other aspects of morphology. In humans, the medial and lateral menisci are crescent-shaped and each attaches to the tibia via two ligaments, one ligament for the posterior horn of the meniscus and a separate ligament for the anterior horn (Tardieu, 1986). In chimpanzees and gorillas, a single ligament attaches a ring-shaped lateral meniscus to the tibia. The single attachment allows the meniscus to translate more freely in an anteroposterior direction which facilitates significant external and internal rotation of the tibia at the knee (Tardieu, 1986). The medial meniscus of the non-human apes has two attachment points, as modern humans do, making it relatively immobile.

Coupled with the difference in lateral meniscus attachment is a difference in the anteroposterior curvature of the lateral tibial condyle. Humans are characterized by relatively flat lateral tibial condyles, whereas the chimpanzee and gorilla lateral tibial condyles are convexly curved (Thompson, 1889; Trinkaus, 1975; Tardieu, 1983). Organ and Ward (2006) argued that the convex curvature of chimpanzee and gorilla lateral tibial condyles provides greater stability throughout the range of knee flexion-extension. The convex lateral tibial condyles of chimpanzees and gorillas, paired with the mobile lateral meniscus, are likely related to their greater capacity for internal and external rotation at the knee. Tardieu (1986) demonstrated via manipulation of cadaveric limbs that chimpanzees have a much larger range of internal/external rotation at the knee (∼30–40 degrees more). Research on human and chimpanzee knees indicates that this rotation occurs about a longitudinal axis that runs through the medial compartment of the knee parallel to the shaft of the tibia (Tardieu, 1986; Freeman and Pinskerova, 2005).

Although both the distal femur and aspects of the lateral tibial condyle have been scrutinized for functional and behavioral differences between humans and non-human apes, the medial tibial condyle has received less attention. Sylvester and Organ (2010) examined the anteroposterior and mediolateral curvature of the medial tibial condyle, but did not find evidence for differences between humans, chimpanzees, and gorillas or an effect of body mass on curvature. Difference in the range of longitudinal knee rotation and medial femoral condyle shape between human and non-human apes suggests, however, that there may be differences in the shape of the medial tibial condyle. Here I use geometric morphometric techniques to determine whether shape differences exist between the medial tibial condyles of human, chimpanzees, and gorillas. I test the null hypothesis that medial tibial condyles are the same shape in all species and examine any shape differences in light of some of the known differences in knee function and morphology.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

Sample and Data Collection

The sample consists of tibiae from 135 African hominid specimens curated as parts of the Hamann-Todd Osteological Collection (Cleveland Museum of Natural History), the William M. Bass Skeletal Collection (The University of Tennessee), and the Taï chimpanzee skeletal collection (Max Planck Institute for Evolutionary Anthropology) (Table 1). All specimens are free from pathology and skeletally adult. Polyvinylsiloxane molds (President Jet Regular, Coltene-Whaledent) of the proximal tibia of specimens that are part of the Hamann-Todd Collection were prepared (Galbany et al., 2006), and molds were scanned using a NextEngine laser scanner. Tibiae that are part of the Taï chimpanzee skeletal collection were scanned using a Breuckmann optoTop-HE white light surface scanner. Both scanners directly produce triangulated mesh surface models. The white light scanner produced surface models with vertices that were on average less than 0.3 mm apart, and the laser scanner produced models with vertices that were less than 0.2 mm apart. Human tibiae from the William M. Bass Skeletal Collection were scanned on a GE Lightspeed Computed Tomography 16-slice scanner (100 kVp, 150 mA, Filter = “Body Filter”) with isometric cubic voxels (0.625 mm). Image stacks were manually segmented in commercially available software (Avizo®, Visualization Sciences Group, Burlington, MA) resulting in surface models of the bones (Sylvester et al., 2008). All surface models of the proximal tibiae were imported into a virtual workspace (Geomagic Studio®, Geomagic, Inc., Morrisville, NC) and the medial tibial condyle was trimmed from the rest of the bone surface along the articular margin (Fig. 1).

image

Figure 1. Medial tibial condyle trimmed from the rest of the proximal tibia (Pan troglodytes).

Download figure to PowerPoint

Table 1. Sample
SpeciesMalesFemales
Homo sapiens  
Bass1713
Hamann-Todd1014
Pan troglodytes  
Max Planck Institute66
Hamann-Todd1620
Gorilla gorilla  
Hamann-Todd1320

The process of trimming the articular surface from the rest of the tibia was performed two additional times for 12 of the tibiae (two of each sex for each species) over a period of weeks. The geometric morphometric analysis described below was performed on replicates and all Procrustes distances between replicates were found to be smaller than the smallest Procrustes distance between specimens (entire sample). On average, replicate Procrustes distances were smaller by more than an order of magnitude (average Procrustes distance between replicates was 0.008 compared to 0.100 between specimens).

Analysis

To quantify the three-dimensional shape of the medial tibial condyle, 504 sliding semi-landmarks were distributed across each joint surface using custom software written for Matlab® (MathWorks, Inc., Natick, MA) following Gunz et al. (2005). Seventy-one of these points were placed along the articular margin and the remaining 433 were placed on the articular surface (Fig. 2). All landmarks were slid iteratively along tangent planes (surface landmarks) and curves (articular margin landmarks) to minimize the bending energy of the thin-plate spline interpolation function between each specimen and the updated Procrustes average (Gunz et al., 2009). I converted specimens to shape coordinates by carrying out generalized Procrustes superimposition, which removes information about location and orientation and scales all specimens by centroid size (Rohlf and Slice, 1990). Shape variation was summarized using principal component analysis and average medial condyles were created for each species and sex within species.

image

Figure 2. Medial tibial condyle with 504 sliding semi-landmarks (Pan troglodytes).

Download figure to PowerPoint

Randomization tests were used to test for differences between means shapes of species and sexes within species (Sokal and Rohlf, 1995). The randomization tests compared the Procrustes distances between the mean shapes of two groups (i.e., species or sexes) to a distribution of Procrustes distances. This distribution was created by pooling all individuals from the two groups being compared, and then randomly assigning individual specimens to one of two groups that had sample sizes of the original groups. The Procrustes distance between the means of the randomized groups was calculated, and the randomization procedure was carried out 10,000 times for each comparison. This procedure tests whether the difference between the means of the two groups is greater than would be expected by random chance.

To confirm differences in curvature that were found in the geometric morphometric analysis, two angle measurements were collected on average male and female species medial tibial condyle surfaces created from the geometric morphometric analysis. To do this, each average condyle shape was oriented by fitting the main portion (excluding the eminence) to a horizontal plane. An anteroposterior transect across the articular surface was extracted following Sylvester and Organ (2010). Instead of extracting the contour at 50% of mediolateral width as Sylvester and Organ (2010) did, the transect was positioned to pass through the deepest (most inferior) point of the central portion of the articular surface. An angle between two lines was then measured on the contour. The first line passed through the deepest point and the most superior point on the anterior portion of the contour, while the second line passed through the deepest point and the most superior point on the posterior portion (Fig. 3). An analogous process was used to extract a mediolateral transect contour and measure a mediolateral angle. This contour was taken perpendicular to the anteroposterior transect and passed through the most superiorly projecting point on the intercondylar eminence (Fig. 3).

image

Figure 3. Contours extracted to measure anteroposterior and mediolateral angles. A: Anteroposterior transect (black line) was taken through the deepest point (most inferior) of the central portion of the articular surface, and mediolateral transect (black line) was taken perpendicular to the anteroposterior transect through the most superior projection of the intercondylar eminence. B: Anteroposterior angle: Black line represents the anteroposterior contour across the condyle. The two gray lines represent the lines used to measure the angle, running from the most inferior point on the articular surface to the most superior points on the anterior and posterior portions of the condyle. C: Mediolateral angle: Black line represents the mediolateral contour across the codyle. The gray lines are the same as for the anteroposterior angle.

Download figure to PowerPoint

To determine whether body mass influences the shape of the medial tibial condyle, the principal component scores resulting from the geometric morphometric analysis of the entire sample were regressed against superoinferior femoral head diameter (used as a proxy for body mass, Sylvester and Organ, (2010)) within species. None of the regression analyses revealed a statistically significant relationship and as a result are not reported.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

The first two principal components of shape space separate the three species examined here, and randomization tests indicate significant shape differences between species pairs (Fig. 4 and Table 2). Humans are well separated from the other two species along the first principal component, while the separation of the chimpanzees and gorillas along the second component is less dramatic. Together the first two principal components account for ∼44% of the total shape variance. The first principal component, which accounts for 33% percent of the shape variation, describes a shape transition from an oval-shaped medial tibial condyle (found in humans) to a more triangular-shaped medial tibial condyle in the non-human apes (Fig. 5). The second component describes a change in the angle between the main portion of the condyle and the lateral portion of the articular surface that covers the intercondylar eminence (best viewed from either an anterior or posterior perspective) (Fig. 5). In gorillas, the lateral portion of the articular surface that covers the intercondylar eminence rises more steeply away from the main portion of the condyle thus forming a more acute angle between these two portions of the articular surface.

image

Figure 4. Principal component analysis in shape space. PC1 and PC2 explain approximately 44% of the sample variance. The shape differences associated with the principal components are plotted as surface deformations of the mean shape (2 standard deviations in either direction). Shapes along PC1 are superior in view; those along PC2 are from posterior perspective. Open circles = Homo sapiens females; Closed circles = Homo sapiens males; Open diamonds = Pan troglodytes females; Closed diamonds = Pan troglodytes males; Open triangles = Gorilla gorilla females; Closed triangles = Gorilla gorilla males.

Download figure to PowerPoint

image

Figure 5. Average right medial tibial condyle shapes. Images above species names are superior views with anterior at the top of the image and medial to the left. Images below the species names are posterior views with medial to the left of the image.

Download figure to PowerPoint

Table 2. P-Values for Randomization Tests for Differences Between Means
ComparisonProcrustes distance
Human-Chimp<0.001
Human-Gorilla<0.001
Chimp-Gorilla<0.001

Tests for differences in mean shape do not indicate differences between human males and females, but do indicate shape differences between the sexes within chimpanzees and gorillas (Table 3). When male and female average medial tibial condyles are aligned within chimpanzees and gorillas, subtle shape differences can be appreciated that relate to the curvature across the condyle. The male gorilla joint surface appears more curved in both anteroposterior and mediolateral directions as compared to the female joint surface. The central portion of the male gorilla condyle falls below that of the female, while rising above it on the edges of the condyles (Fig. 6). The angle measurements take from the transect profiles on the average male and female condyles confirm both of these trends. Among gorillas, the angle is smaller for the average male in both the mediolateral and anteroposterior directions (133.6 and 168.6 degrees in mediolateral and anteroposterior directions, respectively) compared to the average female (134.9 and 171.3 degrees in mediolateral and anteroposterior directions, respectively). The chimpanzee sexes are nearly identical in their mediolateral angle (male = 141.4 degrees, female = 141.2 degrees), while in the anteroposterior direction the average male angle is 170.5 degrees and the average female angle is 172.0 degrees. Although the randomization test did not indicate statistically significant shape differences between the human sexes, they follow a pattern of the measured angles similar to the gorilla pattern. The human male angles are 138.5 degrees (mediolateral) and 168.1 degrees (anteroposterior), while the human female angles are 141.7 degrees (mediolateral) and 168.6 degrees (anteroposterior).

image

Figure 6. Average right male and female medial tibial condyle shapes. Left column of images: Gorilla gorilla; right column: Pan troglodytes. Dark gray mesh surfaces are male average surfaces and light gray surfaces are female average surfaces. A = anterior; M = medial.

Download figure to PowerPoint

Table 3. P-Values for Randomization Tests for Differences Between Mean Sex Shapes Within Species
ComparisonProcrustes Distance
Human0.4970
Gorilla0.0060
Chimpanzee0.0046

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

Known differences in knee morphology and function between humans, chimpanzees, and gorillas provide potential explanations for the shape differences in the medial tibial condyle observed here. Tardieu (1981, 1983) demonstrated a fundamental shape difference between the distal femora of humans and non-human apes using a ratio of maximum anteroposterior and mediolateral dimensions. In humans, the anteroposterior and mediolateral lengths are nearly equal, while in chimpanzees and gorillas the distal femur has a much smaller anteroposterior dimension compared to its mediolateral width. Tardieu (1986) also demonstrated that chimpanzees (and presumably gorillas) have a much larger range of internal/external rotation at the knee compared to humans across the entire range of knee flexion.

Separate lines of evidence suggest that internal/external rotation at the hominid knee occurs about a longitudinal axis passing through the medial compartment of the knee. In chimpanzees and gorillas, the lateral meniscus of the knee has a single attachment point, allowing it to move more freely in an anteroposterior direction compared to the relatively stable medial meniscus, which has two ligamentous attachment points (Tardieu, 1986). This suggests that the axis of rotation is located in the more stable medial compartment. Churchill et al. (1998) loaded human whole lower limb anatomical samples and tracked the relative motion of the tibia and femur, finding that internal/external rotation indeed occurs about an axis through the medial tibial condyle that is roughly parallel to the anatomical axis of the tibia. More recently, magnetic resonance imaging studies of the human knee have demonstrated that during internal/external rotation, the lateral femoral and tibial condyles undergo much greater relative motion, compared to a stable medial compartment (Freeman and Pinskerova, 2005), again demonstrating that the axis of internal/external rotation passes through the medial tibial condyle.

First Principal Component

The human medial tibial condyle is more oval-shaped and anteroposteriorly elongated while those of the chimpanzee and gorilla are mediolaterally expanded (relative to their anteroposterior dimension) and approach a distinctly triangular shape. This shape difference relative to humans can be explained in terms of providing an adequate area of support for the medial femoral condyle. The human medial femoral condyle is anteroposteriorly elongated (relative to mediolateral width) compared to chimpanzees and gorillas. In the human knee, the medial femoral condyle is supported not only directly by contact with the medial tibial condyle, but also indirectly through medial meniscus (Messner and Gao, 1998). As a result the human medial tibial condyle may need to be similarly elongated in order to provide an area of support below the femoral condyle. The distinctly triangular shape of the chimpanzee and gorilla medial tibial condyle may reflect the need to maintain contact and provide an area of support for the medial femoral condyle through a larger range of internal/external knee rotation.

The chimpanzee knee can accommodate 40 degrees of combined internal and external rotation (Tardieu, 1986). This rotation creates a more obtuse angle between the sagittal plane anteroposterior axes of the medial femoral and tibial condyles. While the curvatures of the medial femoral and tibial condyles are different, resulting in a relatively small area of direct contact between the cartilage of the articular surfaces; the congruity of the surfaces is dramatically increased by the presence of the medial meniscus. The medial meniscus covers ∼60% of the medial condyle in humans (Clark and Ogden, 1983) and transmits up to 50% of the load which passes through the medial compartment (Messner and Gao, 1998). Loss of the meniscus has been shown to decrease the contact area between the femur and tibia by 30–50% (Fukubayashi and Kurosawa et al., 1980; Baratz et al., 1986; Ihn et al., 1993). Thus contact area between the tibial and femoral condyles, including that mediated by the meniscus, is not limited to the central portion of articular surfaces. The triangular shape of the medial tibial condyle in chimpanzees and gorillas likely provides a medial expansion of the condyle necessary to provide support to the meniscus and medial femoral condyle as it rotates relative to the tibia (Fig. 7).

image

Figure 7. The effect of medial femoral condyle shape and internal/external rotation at the knee on medial tibial condyle shape. Upper row = Femoral and tibial knee components in H. sapiens. Lower row = Femoral and tibial knee components in P. troglodytes. Column A: Outlines of distal right femora from an inferior view, with the medial condyles shaded gray. Column B: Idealized outlines of the medial femoral condyles. Column C: Condyle outlines rotated to maximal internal and external rotation following the experimental data in Tardieu (1986). Column D: Shape of medial tibial condyles that would contain the medial femoral condyle throughout the range of external and internal rotation.

Download figure to PowerPoint

Second Principal Component

The second principal component of shape space describes a difference in the angle between the main (horizontal) body of the medial tibial condyle and the portion of the condyle that extends onto the intercondylar eminence. This trend is confirmed by the angle measurement taken on mediolateral transects. The gorilla medial tibial condyle, in which the angle is more acute, is distinguished from a more obtuse angle in human and chimpanzee medial tibial condyles. This difference is possibly related to load transmission through the medial compartment of the knee in conjunction with requirements for internal/external rotation. The medial femoral condyles of gorillas and chimpanzees are mediolaterally wider than the lateral femoral condyles to accommodate the greater load borne by the medial compartment of the knee during quadrupedal walking (Preuschoft, 1971; Preuschoft and Tardieu, 1996). In humans, loads through the medial and lateral femoral condyles are nearly equal and this equal load transmission is reflected in the equal mediolateral dimensions of the human medial and lateral femoral condyles (Preuschoft, 1971). The differential loading in the non-human ape knee, combined with the greater body mass of gorilla, must result in higher loads through the medial compartment as compared to humans. Male chimpanzees, however, approach the body mass of female gorillas (Smith and Jungers, 1997), and yet the female gorilla medial tibial condyle forms a more acute angle relative to the rest of the condyle compared to male chimpanzee. This suggests that body mass is not the only factor affecting the chimpanzee and gorilla morphologies. The more obtuse angle of the chimpanzee knee may reflect a greater capacity for internal/external rotation at the knee associated with greater arboreal behaviors. Doran (1996, 1997) reports a much greater level of arboreality in chimpanzees (33–68% arboreal) compared to gorillas (2–13% arboreal), including more quadrumanous climbing. Remis (1995, 1998) reports higher levels of arboreality among lowland gorilla at the Bai Hokou Study Site, but cautions that these gorillas were not fully habituated and difficult to see on the ground, precluding accurate reconstruction of percent time in the trees. Because the intercondylar eminence rises into the intercondylar notch of the femur, the steep rise of this feature in gorillas may provide greater stability against mediolateral translation (Blackburn and Craig, 1980; Tardieu, 1981) that may occur during some activities, while sacrificing capacity for internal/external rotation. In humans, the stereotypically low mediolateral loading during walking (Chao et al., 1983) and running (Cavanagh and Lafortune, 1980) may not require an eminence that rises as steeply away from the rest of the condyle.

Sex Shape Differences Within Species

Aligning the average male and female gorilla medial tibial condyles demonstrates that the middle portion of the male condyle falls below that of the female while the edge of the male articulation rises above the female. This indicates that the medial tibial condyle of the male gorilla is more concave than that of the female, a finding confirmed by the angle measurements. A similar pattern can also be seen in the chimpanzee medial tibial condyle, but mainly in an anteroposterior direction. The anterior and posterior edges of the male chimpanzee articular surface rise above the female chimpanzee articular surface, while the middle portion of the male chimpanzee articular surface dips below the female surface. This suggests a greater degree of curvature in the anteroposterior direction. The flatter surfaces of the female medial tibial condyles, compared to their male counterparts, may reflect both differences in body mass as well as differences in locomotor behavior. In both chimpanzees and gorillas, females are both more arboreal (Doran, 1996; Remis, 1999) and smaller than males (Smith and Jungers, 1997). While it is not possible to tease apart the relative contribution of locomotor and body size dimorphism on the shape dimorphism of the gorilla and chimpanzee medial tibial condyles, it is perhaps informative that the human sexes do not display statistically significant differences in shape. Human body size dimorphism approaches that of chimpanzees (Smith and Jungers, 1997), but differences between the human sexes in knee kinematics during locomotion have not been found consistently (Malinzak et al., 2001; Ferber et al., 2003). This suggests that differences in knee morphology of male and female chimpanzees and gorillas may be driven more by differences in locomotor behavior, particularly the degree of arboreality, and less by differences in body size.

CONCLUSION

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

Here I document shape variation in the medial tibial condyle that distinguishes modern humans, chimpanzees, and gorillas. The major source of shape variation describes the outline (from superior view) of the articular surface. Humans have a more oval-shaped medial tibial condyle, while the non-human apes have a more triangular-shaped condyle. This difference likely reflects the greater capacity for internal/external rotation at the knee in the non-human apes and differences between these species in the shape of the medial femoral condyle. The second component, which distinguishes gorillas from chimpanzees and humans, is interpreted as being a product of greater load transmission in the medial compartment of the knee of gorillas, along with lower arboreality among gorillas compared to chimpanzees. Sex differences in shape within the non-human ape species are not directly attributable to body mass, but instead may indicate differences in the degree of arboreality.

ACKNOWLEDGEMENTS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED

The author would like to thank M. Mahfouz and J. Organ for access to tibia scans, B. Latimer, L. Jellema, L.M. Jantz, and R.L. Jantz for access to specimens in their care, and P.A. Kramer for critiques of earlier versions of this manuscript. The author would also like to thank the three anonymous reviewers for their critiques and comments that improved the work presented here.

LITERATURE CITED

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. ACKNOWLEDGEMENTS
  8. LITERATURE CITED