Australopithecus anamensis: A finite-element approach to studying the functional adaptations of extinct hominins
Article first published online: 3 MAR 2005
Copyright © 2005 Wiley-Liss, Inc.
The Anatomical Record Part A: Discoveries in Molecular, Cellular, and Evolutionary Biology
Special Issue: Finite Element Analysis in Vertebrate Biomechanics
Volume 283A, Issue 2, pages 310–318, April 2005
How to Cite
Macho, G. A., Shimizu, D., Jiang, Y. and Spears, I. R. (2005), Australopithecus anamensis: A finite-element approach to studying the functional adaptations of extinct hominins. Anat. Rec., 283A: 310–318. doi: 10.1002/ar.a.20175
- Issue published online: 15 MAR 2005
- Article first published online: 3 MAR 2005
- Manuscript Accepted: 13 JAN 2005
- Manuscript Received: 12 JAN 2005
- Leverhulme Trust. Grant Number: F/00025/A
- Natural Environment Research Council. Grant Number: NER/A/S/2003/00347
- Australopithecus anamensis;
- enamel microstructure;
- finite-element stress analyses;
Australopithecus anamensis is the stem species of all later hominins and exhibits the suite of characters traditionally associated with hominins, i.e., bipedal locomotion when on the ground, canine reduction, and thick-enameled teeth. The functional consequences of its thick enamel are, however, unclear. Without appropriate structural reinforcement, these thick-enameled teeth may be prone to failure. This article investigates the mechanical behavior of A. anamensis enamel and represents the first in a series that will attempt to determine the functional adaptations of hominin teeth. First, the microstructural arrangement of enamel prisms in A. anamensis teeth was reconstructed using recently developed software and was compared with that of extant hominoids. Second, a finite-element model of a block of enamel containing one cycle of prism deviation was reconstructed for Homo, Pan, Gorilla, and A. anamensis and the behavior of these tissues under compressive stress was determined. Despite similarities in enamel microstructure between A. anamensis and the African great apes, the structural arrangement of prismatic enamel in A. anamensis appears to be more effective in load dissipation under these compressive loads. The findings may imply that this hominin species was well adapted to puncture crushing and are in some respects contrary to expectations based on macromorphology of teeth. Taking together, information obtained from both finite-element analyses and dental macroanatomy leads us to suggest that A. anamensis was probably adapted for habitually consuming a hard-tough diet. However, additional tests are needed to understand the functional adaptations of A. anamensis teeth fully. © 2005 Wiley-Liss, Inc.
The oldest hominins, Orrorin tugenensis, Sahelanthropus tchadensis, and Ardipithecus kadabba, are contentious and both their taxonomic status and phylogenetic relationships are hotly debated (Balter, 2001; Senut et al., 2001; Brunet et al., 2002; Wolpoff et al., 2002; Haile-Selassie et al., 2004). Uncertainties stem from the fragmentary nature of the material and the paucity of comparable parts among them, while the polarity of the characters traditionally associated with hominins (i.e., bipedality, canine reduction, and thick-enameled teeth) has recently come under scrutiny also. As a case in point, while enamel thickness is generally considered a defining feature of hominins, a recent study emphasizes the importance of canine reduction and morphology over this character (Haile-Selassie et al., 2004). Less contentious with regard to its phylogenetic position is A. anamensis at about 4.2–3.9 Ma from well-dated deposits at Kanapoi and Allia Bay east of Lake Turkana (Leakey et al., 1995, 1998). This species shows clear adaptations toward bipedalism, canine reduction, and thick-enameled teeth (Leakey et al., 1995, 1998; Ward et al., 1999, 2001) and is generally regarded the stem species of all later hominins. While the postcranium clearly indicates that the species was bipedal when on the ground (Leakey et al., 1995, 1998; Ward et al., 1999, 2001), the functional consequences of the species' masticatory apparatus remain inconclusive (Ward et al., 1999, 2001). The cranial features are primitive and in many respects resemble those of extant great apes (Ward et al., 1999, 2001), but preliminary analyses, primarily based on overall tooth size, enamel thickness, and mandibular corpus size, led to propositions that A. anamensis may have exploited a dietary niche different from the extant great apes as well as from later hominins (Teaford and Ungar, 2000; Ward et al., 2001). Although thick-enameled (and larger) teeth are indeed traditionally associated with hard object feeding (Kay, 1981), a theoretical study has shown that such inferences may not necessarily be warranted (Macho and Spears, 1999). Without structural reinforcement of its internal structure, thick enamel, because of its anisotropic structure, is brittle and prone to failure. Strength and fracture resistance are conferred to the tissue by differently oriented bundles of prisms (i.e., decussation) and by systematic differences in crystal orientation between the prism heads and the interprismatic matrix (von Koenigswald et al., 1987; Rensberger, 2000). This overall correlation between the tissue's ability to absorb load and its microstructural features is undisputed, but the precise relationships are understood only poorly. Biomechanical tests are destructive and although they provide information about the bulk behavior of the tissue, they are unable to pinpoint areas of weaknesses within the tissue (Popowics et al., 2004). As these areas of weaknesses occur on a microstructural level (Rasmussen et al., 1976), they are inaccessible to traditional stress/strain measurements (e.g., strain gauges). Here we present a new approach to overcome these difficulties. By creating finite-element models of virtual test specimens of deviating (and decussating) enamel of A. anamensis teeth and by subjecting these models to compressive stress, we inquire (nondestructively) whether the enamel of A. anamensis may have been adapted to cope with high levels of stress. In the present study, the loads applied to the enamel blocks (i.e., perpendicular to the predominant long axes of the prisms) are not directly related to certain masticatory functions, but are chosen to test propositions that the microstructural arrangement may (or may not) confer strength to the dental tissue irrespective of enamel thickness. Inferences about the dietary niche of A. anamensis can thus be based on more informed evidence.
MATERIALS AND METHODS
Prior to creating the finite-element models, reconstruction of the enamel microstructure from naturally broken surfaces was undertaken using recently developed software (Jiang et al., 2003; Macho et al., 2003). Briefly, the mathematical algorithms underlying the graphic model are based on the assumption that prism deviation (and consequently decussation) among primates is brought about by biophysical processes, i.e., the interplay of secretion of ameloblasts and the cell-cell adhesion among them. The imbalance of forces thus created at the advancing enamel front will cause prisms to buckle. However, the contribution of each of these factors will change over time, i.e., throughout the lifetime of an ameloblast and from earlier- to later-forming enamel. Validation of the model's predictive capabilities against measured primate specimens using controlled breaks further revealed that there are systematic differences in prism arrangements between even closely related species (Jiang et al., 2003; Macho et al., 2003). Because of the interconnectivity of the advancing enamel front, however, differences in prism organization within a tooth (species) are of degree rather than kind (Macho et al., 2003). This is apparent even on a macroscopic level, e.g., when inspecting longitudinal breaks (Fig. 1). In general, the species-specific pattern of undulation is most strongly expressed toward the cusp, where enamel is thickest, and least toward the cervical margin, where prisms tend to be relatively straight. Preliminary analyses also indicate that functional cusps exhibit a higher degree of prism decussation (particularly through reduced cycle length), although the general pattern of undulation is comparable to that of guiding cusps (data not shown). These observations may explain the tight, statistically significant correlations between prism undulation and enamel thickness within and between species (Fig. 3). Differences in prism arrangement can thus be successfully exploited for paleobiological and functional enquiries.
Sample Preparation and Graphic Modeling
Naturally broken enamel surfaces of the following A. anamensis specimens were lightly etched for 20 sec with 5% HCl and high-resolution casts (PROVIL light, epoxy resin) were taken for SEM (backscatter) analyses. The specimens studied are KNM-KP 29287F (rM3), KNM-KP 29287G (lM3), KNM-KP 31715A (lM1/2), KNM-KP 31717C (lM2), KNM-KP 31721A (rM2), KNM-KP 31721B (lM3) KNM-KP 31732B (?), KNM-KP 31732B(2) (lM3), KNM-KP 34725F (rI2), KNM-KP 34725J (rI2), KNM-KP 34725N (lLC), KNM-KP 35839C (lP3), and KNM-KP 35851 (lM2/3). Not all impressions revealed clear microstructures and the breaks were not always along clearly defined planes with regard to the tooth axes. Thus, modeling (Jiang et al., 2003; Macho et al., 2003) was restricted to only four specimens, but inspection of the other specimens revealed comparable patterns of prism decussation (Fig. 2). Unfortunately, only a single transverse break was available for study (KNM-KP 29287F), but this was not very informative (i.e., broken obliquely and too far cervically). Consequently, the predictions of prism deviation in the tangential plane must be considered preliminary. Given the nature of the sample, error margins for the reconstructed enamel microstructures of A. anamensis are likely to be greater than they are for the extant species (Jiang et al., 2003; Macho et al., 2003). Despite this, however, the virtual breaks induced in the graphic models compare well visually with the appearance of real broken enamel surfaces (Fig. 2). Morphological comparisons were made with previously published data (Jiang et al., 2003; Macho et al., 2003) (Figs. 2 and 3).
Finite-Element Modeling and Analyses
To determine the biomechanical behavior of the different enamels, the graphic models of decussating enamel from comparable regions of the guiding cusps of A. anamensis, Pan troglodytes, Gorilla gorilla, and Homo sapiens were converted to composite finite-element models. The geometry of the three-dimensional models, taken from the (mid)crown area of guiding cusps, was imported into MSC.Mentat, the finite-element preprocessing software (MSC Software, 2002). This process involved recreating prism cross-sections (x-y plane) and extruding the section in the out-of-plane dimension. The controlling geometry for the extrusion was a square B-spline imported from the graphic models (Fig. 2). Each model was then expanded to create a cuboid enamel block encompassing at least one full cycle of deviating prisms in the z-y plane for Pan (M1, 126 μm:140 μm:695 μm), Gorilla (M1, 125 μm:139.5 μm:685 μm), Homo (M3, 246 μm:270 μm:1,335 μm), and KNM-KP 35851 (M2/3, 232.2 μm:256 μm:1,270 μm); the longest dimension of each specimen represents the respective enamel thickness from the dentinoenamel junction (DEJ) to the outer enamel surface (OES) along the long axes of the predominant direction of the prisms. The dimensions of the enamel blocks were roughly proportional (i.e., Pan = 1:1.10:5.52; Gorilla = 1:1.12:5.48; Homo = 1:1.10:5.43; A. anamensis = 1:1.10:5.47).
Each prism path was divided into 28 elements. The length of these elements was based on the local curvature of the controlling splines. Hence, elements in regions of high prism deviations are shorter in length than those in regions of low deviations (Fig. 2). In cross-section, each prism consists of four elements (Fig. 4, Table 1). Following Spears (1997), different properties were assigned to each element to take into account differences in crystal orientation (Fig. 4, Table 1).
|Dimensions of elements for each prism (μm)|
|Crystal orientation within each element (degrees)|
|Young's modulus (E/GPa)||32.1||32.1||109|
|Shear modulus (G/GPa)||13.6||31||31|
|Poisson's ratio (ν)||0.177||0.0976||0.313|
On a microstructural level, enamel is made up of a complex arrangement of enamel prisms, whereas on an ultrastructural level, enamel is composed of differently oriented hydroxyapatite crystals held together by an inorganic matrix. Given that the crystals are considerably stiffer than the matrix, enamel (as most biological materials) behaves in a complex manner. Specifically, in cases where loads are applied along the direction of the crystals, most of the internal stresses are carried by the crystals and hence the behavior of enamel is similar to that of crystals (i.e., higher stiffness). In contrast, when loads are applied across the direction of crystals, most of the internal stresses are carried by the inorganic matrix and the behavior of enamel in this direction is more similar to that of the matrix. Consequently, the stiffness of enamel is different in different directions, i.e., it is anisotropic with respect to stiffness. For modeling purposes, this behavior of enamel can be simplified by assuming that the crystals are oriented in parallel within a small region, i.e., within one element (Fig. 4). This allows enamel to be modeled with orthotropic (i.e., a simple representation of anisotropy) behavior, in which the material properties are defined along three directions [x-, y-, and z-axes; Fig. 4, Table 1; see also Shimizu et al. (2005) for more details]. The orientation of crystals within each element was defined following Waters (1980) (Fig. 4) and the local properties were calculated using equations based on composite theory (Fung, 1977).
For validation, a small piece of straight enamel was created and its biomechanical behavior (i.e., Young's moduli) was appraised against published experimental data (Craig et al., 1961; Stanford et al., 1960; Xu et al., 1998). When comparing the data in Figure 4, it needs to be borne in mind that these experimental studies either did not specify the loading direction (Stanford et al., 1960) or did not test the tissue's behavior in the y-direction. However, as regards the tissue's behavior in the x- and z-direction, the finite-element results obtained in our validation experiments compare well with those derived from experimental studies; using paired t-tests, the results are statistically significant at the 0.5% probability level (Fig. 4). With these results being satisfactory, each model of decussating enamel was expanded by 20% to add a dentine block at the DEJ, with an isotropic Young's modulus of 16.6 GPa (Macho and Spears, 1999). The total number of elements for each model is thus Pan = 257,869; Gorilla = 168,784; Homo = 288,601; A. anamensis = 272,503. These models were then subjected to an applied pressure of 1 MPa as predicted to occur during human mastication (Fernandes et al., 2003), which was applied perpendicular to the predominant direction (i.e., y-direction) of the enamel block. The model was fixed inferiorly at the x-z plane. Due to lateral deflection induced by the load employed and microstructural inhomogeneities, tensile stresses across the prisms occurred. Such internal tension arises under compressive loads that would occur during mastication and are potentially harmful to the structure of enamel (Rensberger, 2000) and the relative build-up of these stresses is therefore reported for comparative purposes (Figs. 5–7).
Limitations of Finite-Element Models
It should be noted that there are several limitations with regard to the finite-element models and the validation process, which may or may not be overcome in the future. With regard to the models, it is assumed that the crystal orientation between prism head and interprismatic matrix (IPM) and the chemical composition of the enamel matrix are the same for all species and throughout the tissue (Cuy et al., 2002). Also, the structural detail of the simulated dentine and the DEJ (Marshall et al., 2001, 2003) is simplistic when compared to that of enamel. However, preliminary analyses indicate that due to the relatively low stiffness and direction of loading, detailed modeling of the structure of dentine and the DEJ does not affect stress distribution within the enamel. Hence, an isotropic Young's modulus (E = 16.6 GPa) was considered sufficient to represent this behavior (Spears and Macho, 1998). A more important limitation of the model in terms of the calculations of stress is the constraints assigned along the x-z plane. Ideally, in order to overcome this limitation, the whole tooth (or at least a larger piece of enamel) would have to be modeled. However, given the level of detail used in our microstructural modeling (together with current computational limits), this is not possible at present.
With regard to the validation process, it is noteworthy that although the overall deformation behavior of the model compares well with that in experiments, data do not exist (and cannot be obtained through traditional biomechanical testing, e.g., strain gauges) with which to compare the magnitudes of internal stress (i.e., at the ultrastructural level). In other words, as the localization of stress occurs on a prismatic level, strain gauges would have to be smaller than the dimensions of the prisms in order to measure stresses.
With regard to the loading conditions prescribed, only compressive stress was applied to the enamel blocks. Depending on the diet and stage of the chewing cycle, the direction and position of external loads on the teeth will vary. In this first article, it is hoped that blocks would be tested under the type of stress that would mainly occur when guiding cusps are subjected to vertical cusp-tip loads. However, during later phases of chewing, the functional cusps will undergo loading and the guiding cusps may become loaded laterally and may be subjected to bending. Consequently, the compressive loads applied in the present study do not represent the range of loads occurring during mastication. Also, the static nature of the model is restricting its use to predict fracture initiation. Therefore, the possibility that these localized cracks propagate through the structure and may threaten the integrity of the animal is based on our subjective interpretation of localized stresses.
Taken together, the predictions of actual values of stress, although based on well-proven algorithms, should remain theoretical. However, it should also be noted that by maintaining consistency in all aspects other than prism orientation, the relative magnitudes and locations of stress are suitable for comparative purposes, although caution should be adopted when making inferences about functional adaptations.
In the transverse plane (x-z plane) and in comparable regions within the tooth, prism deviation in A. anamensis appears to be comparable to that in Pan, but in the longitudinal (y-z) plane, A. anamensis more closely resembles Gorilla (Fig. 2, Table 2). The similarities in the longitudinal plane are in part brought about by undulations of prisms in a tangential plane, i.e., in the x-y plane close to the DEJ; this is similar between A. anamensis and gorillas. In terms of enamel thickness, A. anamensis is close to, or exceeds, modern humans in the regions studied (Table 2). Yet relative prism deviation is relatively low in A. anamensis, such that the overall scaling between enamel thickness and true prism length follows the relationship found in the great apes, rather than the thick-enameled humans (Fig. 3). This may indicate relatively little decussation in this species, which may render the tissue susceptible to fracture. To test these propositions, finite-element models of virtual enamel specimens from comparable regions of the guiding cusps were created and compressed along the y-direction, i.e., perpendicular to the direction of greatest stiffness (Shimizu et al., 2005), where potentially damaging tensile stresses across prisms would be expected to be highest (Rasmussen et al., 1976). Differences in prism attitude between species, which would affect the behavior of the tissue during mastication, were not taken into account. Consequently, the models presented here and the inferences drawn should be regarded as preliminary accounts of the strength of these different tissues (Fig. 5).
|Frequency (x-z plane)||Homo/Gorilla > Pan/A. anamensis|
|Amplitude (x-z plane)||Homo > Pan/A. anamensis > Gorilla|
|Frequency (y-z plane)||Gorilla/A. anamensis > Pan > Homo|
|Amplitude (y-z plane)||Homo > Pan > Gorilla/A. anamensis|
|Enamel thickness||A. anamensis > Homo > Pan > Gorilla|
|Distribution across length (x-z)||Pan/Gorilla/A. anamensis > Homo|
|Prism path relative to enamel thickness|
|Frequency (x-z plane)||Gorilla > Homo > Pan > A. anamensis|
|Amplitude (x-z plane)||Homo > Pan > A. anamensis > Gorilla|
|Frequency (y-z plane)||Gorilla > Pan > A. anamensis > Homo|
|Amplitude (y-z plane)||Homo > Pan > Gorilla > A. anamensis|
|Distribution across length (x-z)||Pan/Gorilla/A. anamensis > Homo|
|Possible tangential curves||A. anamensis (?) > Gorilla > Pan > Homo|
The maximum principal stresses yielded in the analyses are comparable among species, but there are differences in location and relative distribution within the tissue (Figs. 5 and 6). More importantly, however, enamel is particularly susceptible to fracture when internal tension develops between prisms with the stress acting across their orientation (Rasmussen et al., 1976); this would result in prisms being torn apart. The magnitudes of such tensile stresses perpendicular the long axes of prisms (i.e., those with the greatest damage potential and which are also perpendicular to the direction of load) are considerably lower in A. anamensis and Gorilla than they are in either Homo or Pan (Fig. 7) and they are also relatively localized close to the DEJ.
Features to be expected in the earliest human ancestors are a bipedal mode of locomotion when on the ground, canine reduction, and thick-enameled teeth (Cela-Condo and Ayala, 2003). Changes in locomotor capabilities arguably have profound implications for the animal's behavior, but shifts in dietary capabilities may be of similar or even greater importance, especially where later species of Australopithecus are concerned (Teaford and Ungar, 2000). Changes in masticatory performance not only allowed early hominins to exploit varied food sources commonly associated with increased climatic fluctuations (Teaford and Ungar, 2000), but to partition the environment among sympatric species.
Teeth are most commonly preserved in the fossil record and, given their direct involvement in the breakdown of food, are the structures on which inferences about dietary adaptations are usually based. Perhaps even more importantly, teeth concentrate stress and must withstand the forces created by the muscles of mastication; they may thus provide a more reliable signal about the masticatory performance than the bony structures overlying them. Regardless, tooth size, shape, and enamel thickness contain both phylogenetic and functional information (Janis and Fortelius, 1988). For example, with regard to the thick enamel of hominins, both the polarity of enamel thickness (Haile-Selassie et al., 2004) and its functional consequences remain unresolved (Macho and Spears, 1999; Ward et al., 1999, 2001; Wood and Strait. 2004), thus making it difficult to address paleobiological questions. To contribute toward resolving these issues, enamel microstructure in A. anamensis teeth was reconstructed and the virtual models thus created were then subjected to biomechanical tests using finite-element stress analyses.
Enamel microstructure of A. anamensis teeth shows a unique combination of features not seen in any of the extant hominoids studied (Fig. 2, Table 2). Despite its thick enamel, however, prism deviation in this hominin is relatively low, resulting in the scaling relationship between projected prism length (i.e., enamel thickness) and true prism length being the same as in the great apes (Fig. 3). Although it could be argued from a phylogenetic perspective that the similarities in enamels between A. anamensis and the African apes may have been expected, such scaling relationships are surprising from a functional perspective: the results could imply only moderate crossing-over of bundles of prisms, i.e., little decussation, and hence low resistance to crack propagation. However, given the complex manner in which bundles of prisms are arranged in A. anamensis, this need not be the case.
Under the same applied compressive stress, the magnitudes of maximum tensile stress are comparable among species (Figs. 5 and 6), but the magnitude and location of stress concentration acting across prisms differ markedly between taxa (Fig. 7). As regards the latter, overall stress concentration is lowest in A. anamensis and Gorilla and is concentrated close to the DEJ, especially in A. anamensis. Given the large degree of decussation in this region as well as the region's close proximity to the crack-preventing mechanism of the DEJ (Marshall et al., 2001, 2003), it is improbable that any cracks initiating in this part will propagate through the enamel structure. In contrast, Pan and Homo have the highest tensile stresses away from the DEJ (i.e., in regions of lower decussation) and consequently may be more prone to fracture. Cracks initiating at these sites could travel easily through the outer enamel part where prisms are relatively straight, especially in modern humans (Jiang et al., 2003). These results make evident that prism deviation per se is only a poor predictor of the biomechanical behavior of the tissue, whereas the complex three-dimensional arrangement of prisms with regard to the direction of load appears to be more informative.
Despite similarities in enamel microstructure and overall magnitude and concentration of stresses in A. anamensis and Gorilla, there are fundamental differences between the two species, which may aid in the reconstruction of the dietary adaptations of A. anamensis. Compared with chimpanzees, gorillas are adapted to a more fibrous, relatively tough and varied diet (Kuroda et al., 1996), with their high-cusped, thin-enameled molars providing sufficient shearing crests to cope with such a fibrous diet (Kay, 1977). Conversely, A. anamensis teeth are thick-enameled, low-cusped, and show a greater degree of asymmetry than those of the great apes (Ward et al., 1999). The results of the present analyses indicate that under the loads employed (i.e., compressive stress), the enamel of A. anamensis is apparently well adapted to cope with high loads acting predominantly across the long axes of prisms. Other factors being equal, it could be argued that this hominin species presumably coped with high loads occurring during phase 1 of the chewing cycle, i.e., puncture crushing, thus supporting propositions that A. anamensis was adapted to a hard-brittle diet. However, such inferences may be premature in light of the fact that the enamel blocks were not tested under all types of stress occurring during mastication (e.g., shear stress). Perhaps even more important, they are also rendered questionable when other factors are considered. As a case in point, the greater asymmetry of teeth when compared to the great apes (Ward et al., 1999) would suggest a greater lateral excursion of the mandible (Spears and Macho, 1998; Macho and Spears, 1999) in this species, which would be required when breaking down tough foods. Hence, when the evidence is considered together, it would seem most parsimonious at present to propose that A. anamensis habitually consumed hard-tough (rather than hard-brittle) foods, while the thick enamel in A. anamensis may have been an adaptation toward wear resistance.
To summarize, although steps are taken to reconstruct the enamel microstructure both accurately and repeatedly, and to subject these virtual models to well-proven tests used in engineering, the functional interpretations must still be regarded preliminary. However, the results yielded from such analyses in conjunction with information derived from the macromechanics of the teeth allow more convincing inferences regarding the habitual diet of extinct hominins.
The authors thank the Trustees of the National Museum of Kenya. Dr. Meave Leakey generously supported the study, provided access to the material, and made useful comments on the manuscript. Thanks to Dr. Callum Ross for the invitation to contribute this manuscript.
- 2001. Scientists spar over claims of earliest human ancestor. Science 291: 1460–1461. .
- 2002. A new hominid from the Upper Miocene of Chad, Central Africa. Nature 418: 145–151. , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , .
- 2003. Genera of the human lineage. Proc Natl Acad Sci USA 100: 7684–7689. , .
- 1961. Compressive properties of enamel, dental cements and gold. J Dent Res 40: 936–945. , , .
- 2002, Nanoindentation mapping of the mechanical properties of human molar tooth enamel. Arch Oral Biol 47: 281–291. , , , , .
- 2003. A novel sensor for the bite force determinations. Dent Mat 19: 118–126. , , , .
- 1977. A first course in continuum mechanics, 2nd ed. Englewood Cliff, NJ: Prentice-Hall. .
- 2004. Late Miocene teeth from Middle Awash, Ethiopia, and early hominid dental evolution. Science 303: 1503–1505. , , .
- 1988. On the means whereby mammals achieve increased functional durability of their dentitions, with special reference to limiting factors. Biol Rev Camb Phil Soc 63: 197–230. , .
- 2003. The three-dimensional visualisation model of prismatic enamel: an aid to interpreting fractured surfaces on a micro-structural level. Arch Oral Biol 48: 449–457. , , .
- 1977. The evolution of molar occlusion in the Cercopithecidae and early Catarrhines. Am J Phys Anthrop 46: 327–352. .
- 1981. The nut-cracker: a new theory of the adaptations of the Ramapithecinae. Am J Phys Anthrop 55: 141–151. .
- 1996. Sympatric chimpanzees and gorillas in the Ndoki Forest, Congo. In: McGrewWC, MarchantLF, NishidaT, editors. Great ape societies. Cambridge: Cambridge University Press. p 71–81. , , , .
- 1995. New four-million-year-old hominid species from Kanapoi and Allia Bay, Kenya. Nature 376: 565–571. , , , .
- 1998. New specimens and confirmation of an early age of Australopithecus anamensis. Nature 393: 62–66. , , , , .
- 1999. The effects of loading on the biomechanical behaviour of molars of Homo, Pan and Pongo. Am J Phys Anthrop 109: 211–227. , .
- 2003. Enamel microstructure: a truly three-dimensional structure. J Hum Evol 45: 821–830. , , .
- 2001. Mechanical properties of the dentinoenamel junction: AFM studies of nanohardness, elastic modulus, and fracture. J Biomed Mat Res 54: 87–95. , , , , .
- 2003. The dentin-enamel junction: a natural, multilevel interface. J Eur Ceramic Soc 23: 2897–2904. , , , , , .
- 2004. Enamel microstructure and microstrain in the fracture of human and pig molar cusps. Arch Oral Biol 49: 595–605. , , .
- 1976. Fracture properties of human enamel and dentin. J Dent Res 55: 1362–1368. , , , .
- 2000. Pathways to functional differentiation in mammalian enamel. In: TeafordMF, SmithMM, FergusonMWJ, editors. Development, function and evolution of teeth. Cambridge: Cambridge University Press. p 252–268. .
- 2005. Effect of prism orientation and loading direction on contact stresses in prismatic enamel of primates: implications for interpreting wear patterns. Am J Phys Anthrop (in press). , , .
- 2001. First hominid from the Miocene (Lukeino Formation, Kenya). C R Acad Sci 332: 137–144. , , , , , .
- 1997. A three-dimensional finite-element model of prismatic enamel: a re-appraisal of the data on the Young's Modulus of enamel. J Dent Res 76: 1690–1697. .
- 1998. Biomechanical behaviour of modern human molars: implications for interpreting the fossil record. Am J Phys Anthrop 106: 467–482. , .
- 1960. Compressive properties of hard tooth tissues and some restorative materials. J Am Dent Assoc 60: 746–756. , , , .
- 2000. Diet and the evolution of the earliest human ancestor. Proc Natl Acad Sci USA 97: 13506–13511. , .
- 1987. Changes in the tooth enamel of early Palaeocene mammals allowing dietary diversity. Nature 328: 150–152. , , .
- 1999. The new hominid species Australopithecus anamensis. Evol Anthrop 7: 197–205. , , .
- 2001. Morphology of Australopithecus anamensis from Kanapoi and Allia Bay, Kenya. J Hum Evol 41: 255–368. , , .
- 1980. Some mechanical and physical properties of teeth. In: VincentJVF, CurryJD, editors. The mechanical properties of biological materials. Cambridge: Cambridge University Press. p 99–135. .
- 2002. Sahelanthropus or “Sahelpithecus.” Nature 419: 581–582. , , , .
- 2004. Patterns of resource use in early Homo and Paranthropus. J Hum Evol 46: 119–162. , .
- 1998. Indentation damage and mechanical properties of human enamel and dentin. J Dent Res 77: 472–480. , , , , , , .