The game of models: Dietary reconstruction in human evolution.

Despite substantial additions to the paleontological record and unanticipated improvements in analytical techniques since the Journal of Human Evolution was first published, consensus on the diet of early hominin species remains elusive. For instance, the notable advances in the analyses of hominin dental microwear and stable isotopes have provided a plethora of data that have in some instances clouded what was once ostensibly a clear picture of dietary differentiation between and within hominin taxa. In the present study, we explore the reasons why the retrodiction of diet in human evolution has proven vexing over the last half century from the perspective of both ecological and functional-mechanical models. Such models continue to be indispensable for paleobiological reconstructions, but they often contain rigid or unstated assumptions about how primary paleontological data, such as fossils and their geological and taphonomic contexts, allow unambiguous insight into the evolutionary processes that produced them. In theoretical discussions of paleobiology, it has long been recognized that a mapping function of morphology to adaptation is not one-to-one, in the sense that a particular trait cannot necessarily be attributed to a specific selective pressure and/or behavior. This article explores how the intrinsic variability within biological systems has often been underappreciated in paleoanthropological research. For instance, to claim that derived anatomical traits represent adaptations related to stereotypical behaviors largely ignores the importance of biological roles (i.e., how anatomical traits function in the environment), a concept that depends on behavioral flexibility for its potency. Similarly, in the paleoecological context, the underrepresentation of variability within the 'edible landscapes' our hominin ancestors occupied has inhibited an adequate appreciation of early hominin dietary flexibility. Incorporating the reality of variation at organismal and ecological scales makes the practice of paleobiological reconstruction more challenging, but in return, allows for a better appreciation of the evolutionary possibilities that were open to early hominins.

Oral processing, sexual selection, and size variation in the circumorbital region of Colobus and Piliocolobus.

OBJECTIVES: The function of the browridge in primates is a subject of enduring debate. Early studies argued for a role in resisting masticatory stresses, but recent studies have suggested sexual signaling as a biological role. We tested associations between circumorbital form, diet, oral processing, and social behavior in two species of colobus monkey-the king colobus (Colobus polykomos) and western red or bay colobus (Piliocolobus badius). MATERIALS AND METHODS: We quantified circumorbital size and dimorphism in a sample of 98 crania. Controlling for age and facial size, we tested whether variation in circumorbital morphology can be explained by variation in diet, oral processing behavior, masticatory muscle size, and mating system. To contextualize our results, we included a broader sample of facial dimorphism for 67 anthropoid species. RESULTS: Greater circumorbital thickness is unrelated to the stresses of food processing. King colobus engages in longer bouts of anterior tooth use, chews more per ingestive event, and processes a tougher diet, yet circumorbital thickness of C. polykomos is reduced compared to P. badius. Differences in circumorbital development do not vary with wear or facial size. Greater sexual dimorphism is present in P. badius; comparisons across anthropoids indicated patterns of circumorbital dimorphism were decoupled from overall size dimorphism. CONCLUSIONS: The expanded circumorbits of male red colobus monkeys evolved in response to intense male-male competition. This hypothesis is consistent with the pattern across anthropoid primates and highlights the underappreciated role of sexual selection in shaping the primate face.

Diet, feeding behavior, and jaw architecture of Taï monkeys: Congruence and chaos in the realm of functional morphology.

We review feeding and mandibular anatomy in a community of West African monkeys. We use field observations, food material property data, and skeletal specimens from the Ivory Coast's Taï Forest to explore the factors that shape mandibular architecture in colobines and cercopithecines. Despite excellent geographic control across our sample, the fit between bone form (as conventionally described) and functional activity (as we perceive it) is not spectacular. We present a thought experiment to assess how well we could reconstruct diet in the Taï monkeys if we only had skeletons and teeth to study. This exercise indicated that we would be correct about half the time. Our analyses reinforce the notion that diet is anything but a monolithic variable and that better success at relating mandibular form to food must incorporate information on ingestive and processing behavior, geometric and material properties of foods, and both material and structural data on jaws themselves.

Oral Processing Behavior of Diana Monkeys (Cercopithecus diana) in Taï National Park, Côte d'Ivoire.

Primates are hypothesized to "fall back" on challenging-to-process foods when preferred foods are less available. Such dietary shifts may be accompanied by changes in oral processing behavior argued to be selectively important. Here, we examine the oral processing behavior of Diana monkeys (Cercopithecus diana) in the Taï Forest across their dietary breadth, testing relationships among food intake, fruit availability, preference, and oral processing behaviors including those involved in food ingestion and breakdown. We conducted 1,066 focal follows from April 2016 to September 2017 documenting frequencies of incisor, canine, and cheek tooth mastications (i.e., chews) per ingestive action (n = 11,906 feeding events). We used phenological survey and scan sample data collected between 2004 and 2009 to examine dietary preference and food availability. Our analyses show that Diana monkeys processed foods in significantly different ways (H2 = 360.8, p < 0.001), with invertebrates requiring the least oral processing, fruit requiring intermediate amounts, and leaves requiring the most oral processing. There was no relationship between fruit availability and consumption of preferred or nonpreferred fruits (p > 0.05); however, preferred fruits were processed with significantly fewer mastications (i.e., less chewing) than nonpreferred fruits (U = 6,557, p < 0.001). We thus demonstrate that, when preferred foods are scarce, Diana monkeys do not fall back on difficult-to-process foods. Changes in processing profiles occurred throughout the year and not solely when preferred foods were in short supply. Though preferred fruits required less processing than nonpreferred fruits, we found no relationship between fruit preference and fruit availability. Diana monkeys' lack of readily identifiable fallback foods may be attributable to the relatively high tree diversity and productivity of the Taï Forest. We conclude that Diana monkeys engage in resource switching, consuming a relatively easy-to-process diet year-round.

Brief communication: Correlation between elastic modulus and radiographic density in mandibular cortical bone of colobine monkeys.

OBJECTIVES: The relationship between radiographic grayscale and elastic modulus was determined using mandibular cortical bone of colobine monkeys. Quantification of this relationship is critical for establishing absolute measures of structural rigidity of skeletal elements. MATERIALS AND METHODS: We determined the Vickers microhardness values in mandibular bone from two species of African colobine monkeys and related these values to elastic modulus through an empirically determined correlation. We also determined radiographic grayscale values from microcomputed tomographic scans of the mandible in the same regions in which microhardness was sampled. We then correlated modulus to grayscale with a power law relationship. RESULTS: We found that elastic modulus scaled with negative allometry with respect to grayscale with an exponent of 0.77. DISCUSSION: Our results suggest a single exponent can effectively capture the relationship of grayscale to elastic modulus and facilitate development of heterogeneous structural models for use in comparative and computational biomechanical studies.

Spatial variation in mandibular bone elastic modulus and its effect on structural bending stiffness: A test case using the Taï Forest monkeys.

OBJECTIVES: We investigated how heterogeneity in material stiffness affects structural stiffness in the cercopithecid mandibular cortical bone. We assessed (1) whether this effect changes the interpretation of interspecific structural stiffness variation across four primate species, (2) whether the heterogeneity is random, and (3) whether heterogeneity mitigates bending stress in the jaw associated with food processing. MATERIALS AND METHODS: The sample consisted of Taï Forest, Cote d'Ivoire, monkeys: Cercocebus atys, Piliocolobus badius, Colobus polykomos, and Cercopithecus diana. Vickers indentation hardness samples estimated elastic moduli throughout the cortical bone area of each coronal section of postcanine corpus. For each section, we calculated maximum area moment of inertia, Imax (structural mechanical property), under three models of material heterogeneity, as well as spatial autocorrelation statistics (Moran's I, IMORAN ). RESULTS: When the model considered material stiffness variation and spatial patterning, Imax decreased and individual ranks based on structural stiffness changed. Rank changes were not significant across models. All specimens showed positive (nonrandom) spatial autocorrelation. Differences in IMORAN were not significant among species, and there were no discernable patterns of autocorrelation within species. Across species, significant local IMORAN was often attributed to proximity of low moduli in the alveolar process and high moduli in the basal process. DISCUSSION: While our sample did not demonstrate species differences in the degree of spatial autocorrelation of elastic moduli, there may be mechanical effects of heterogeneity (relative strength and rigidity) that do distinguish at the species or subfamilial level (i.e., colobines vs. cercopithecines). The potential connections of heterogeneity to diet and/or taxonomy remain to be discovered.

Feeding and oral processing behaviors of two colobine monkeys in Tai Forest, Ivory Coast.

We collected frequency data on oral processing behaviors during feeding in habituated groups of Western red colobus, Piliocolobus badius, and Western black and white, Colobus polykomos, ranging in the Ivory Coast's Tai National Park. During the sampling period, the diet of red colobus consisted of approximately 75% leaves compared to approximately 47% leaves and buds in black and white colobus. Black and white colobus chewed more frequently per ingestive event than did red colobus. Black and white colobus also employed their anterior teeth much more frequently than did red colobus, a difference attributed to the frequent consumption by C. polykomos of Pentaclethra macrophylla seeds and pods. A material analysis of these food items reveals that both the seed coating and seed flesh are quite soft; however, the pod housing the seeds is very tough. We argue that the pod's toughness, geometry, and fiber orientation collectively result in a food that is very difficult to process, resulting in long handling times and frequent, aggressive use of the incisors. We compare these data with those collected on another Tai primate-the sooty mangabey, Cercocebus atys-and demonstrate that during feeding, both colobine species use their incisors less than the mangabey, but that the cercopithecine chews less than either colobine. Combining data on oral processing behaviors with those on the material properties of items being ingested should lead to more informed interpretations of dentognathic morphology.

Comparative biomechanics of Australopithecus sediba mandibles.

Fossils attributed to Australopithecus sediba are described as having phylogenetic affinities with early Homo to the exclusion of other South African australopiths. With respect to functional anatomy of mastication, one implication of this hypothesis is that A. sediba mandibles should exhibit absolutely and relatively reduced stiffness and strength in comparison to Australopithecus africanus and Paranthropus robustus jaws. Examination of cortical bone distribution in the MH 1 and MH 2 mandibles of A. sediba (evaluated against samples of Pan, early and modern Homo as well as A. africanus and P. robustus) indicate that the A. sediba mandibular corpus was geometrically similar to other South African australopiths. In particular, enhanced torsional rigidity is characteristic of all South African australopiths including A. sediba. These findings are consistent with a hypothesis that masticatory mechanics may have been similar to other australopiths (and distinct from exemplars of early Homo), and as such suggest that A. sediba's mandibles were functionally suited to consume hard and tough objects. Recent mechanical modeling of the A. sediba cranium, however, has been interpreted as indicating that this species was relatively poorly adapted to produce large bite forces and likely experienced relatively modest strains in its facial skeleton. This paradox - that the cranium signals a departure from the australopith morphotype whereas the mandibles conform to a hypodigm of australopith grade - can be resolved, in part, if it is acknowledged that mechanical performance variables offer imperfect insight into what constitutes feeding adaptations.

The enduring puzzle of the human chin.

Although modern humans are considered to be morphologically distinct from other living primates because of our large brains, dexterous hands, and bipedal gait, all of these features are found among extinct hominins. The chin, however, appears to be a uniquely modern human trait. Probably because of the chin's exclusivity, many evolutionary scenarios have been proposed to explain its origins. To date, researchers have developed adaptive hypotheses relating chins to speech, mastication, and sexual selection; still others see it as a structural artifact tangentially related to complex processes involving evolutionary retraction of the midfacial skeleton. Consensus has remained elusive, partly because hypotheses purporting to explain how this feature developed uniquely in modern humans are all fraught with theoretical and/or empirical shortcomings. Here we review a century's worth of chin hypotheses and discuss future research avenues that may provide greater insight into this human peculiarity.

Symphyseal surface strain during in vitro human mandibular wishboning.

OBJECTIVES: This research theoretically models and empirically records symphyseal surface strain during in vitro human mandibular wishboning (lateral transverse bending) in order to test one aspect of the hypothesis that the chin is an adaptive response to masticatory stresses. From a perspective of optimality, three questions were tested: 1) Do human mandibles function as curved beams during wishboning? 2) Is the presence of a chin associated with lower than predicted curved beam effects? 3) Are there relatively low strain gradients on the lingual and labial symphyseal surfaces respectively? MATERIALS AND METHODS: Based on morphometric criteria, theoretical wishboning strains were calculated for five dentate adult human mandibles. The same mandibles were fitted with strain gauges and subjected to simulated wishboning loads. From the empirically-recorded strains, relative strains were calculated by dividing all strains by the absolute lowest strain in a given specimen. The theoretical and empirical results were compared in order to address the three related questions guiding this research. RESULTS: Human mandibles behave as curved beams during wishboning (question 1). Empirical strain measures showed greater disparity both between and within the labial and lingual symphyseal surfaces than the theoretical models predictions (questions 2 and 3). DISCUSSION: Human symphyseal form, with its distinctive chin, is unlikely to be adapted for countering wishboning loads. Chins are associated with larger than expected strain gradients within and between symphyseal surfaces, which runs counter to the optimality criterion typically invoked in assessing trait performance for signs of adaptation. The implications are twofold: 1) wishboning may not, in fact, be a regular feature of human mastication or 2) wishboning may not pose the same structural risks in human jaws as this load does in other anthropoid primates.

Bone remodeling is reduced in high stress regions of the cercopithecoid mandible.

OBJECTIVES: Independent lines of evidence suggest that osteonal bone remodeling is a function of both mechanical (i.e., changes in stress) and non-mechanical (i.e., metabolic needs related to calcium liberation) factors. The degree to which secondary bone reflects mechanical factors, however, is incompletely understood despite the common assumption that the stress environment mediates remodeling activity. Here, we investigate whether there are remodeling differences between regions of primate mandibular bone known to have distinct stress environments. METHODS: Osteon density, osteon fragment density, and relative osteonal area are measured as indicators of remodeling activity from postcanine and symphyseal thin sections of four sympatric monkey species (N = 20 total) from Taï Forest, Côte d'Ivoire: Piliocolobus badius, Colobus polykomos, Cercocebus atys, Cercopithecus diana. Subfamily and regional effects were assessed by two-way ANOVA. RESULTS: Symphyseal bone has lower osteonal density, fragment density and relative osteonal area than postcanine bone in all species, indicating relatively low remodeling activity in symphyseal bone, despite the likelihood of relatively high stresses in its lingual cortex. Subfamily differences in postcanine remodeling are significant in that colobines exhibit greater remodeling than cercopithecines. DISCUSSION: The data suggest that high strains are not requisite for bone remodeling. How the lingual symphysis negotiates a high strain environment without the reparative benefit of remodeling is yet to be elucidated, but the data prompt reevaluation of the relationship of targeted remodeling to stress histories. Differences in osteonal bone density between taxa might reflect feeding or dietary influences on remodeling activity, but sample composition prevents parsing behavioral from other taxon-specific effects.

Microbial osteolysis in an Early Pleistocene hominin (Paranthropus robustus) from Swartkrans, South Africa.

Microbiological degradation is one of the most important factors responsible for the destruction of bone in archaeological contexts. Microscopic focal destruction (MFD) is the most prevalent form of microbial tunneling and is encountered very commonly in human bones from archaeological sites, whereas animal bones from these same sites show significantly better preservation if they were deposited in a fragmentary (e.g., butchered) state. Similarly, most fossils show either no evidence or only minor traces of bacterial osteolysis. These observations and experimental evidence point to an endogenous origin for osteolytic bacteria, suggesting that bone bioerosion could potentially aid in reconstructing early taphonomic events. We here report extensive MFD in the mandibular corpus of a small (presumptive female) individual of the hominin Paranthropus robustus from the Early Pleistocene site of Swartkrans, South Africa. The specimen (SKX 5013) derives in situ from the Member 2 deposit, which is dated to ca. 1.5-1.0 Ma. Examination of sections from the corpus by backscattered electron microscopy reveals numerous small linear longitudinal and budded tunneling cavities, which tend to be concentrated around Haversian canals and are more abundant closer to the endosteal aspect of the section. The taphonomy of Swartkrans has been the subject of intense investigation, and given the possibility that different agents of accumulation may have been responsible for the faunal and hominin fossils in the different members at the site, the observation that a specimen of P. robustus from Member 2 displays significant microbial osteolysis is of potential interest. A study of the prevalence of this process in adequately large samples of the animal bones from these units may yield novel insights and provide refinement of our understanding of their taphonomic histories. Such observations might well reveal differences among the various members that could provide another valuable source of osteoarchaeological information for the site.

Spatial patterning of bone stiffness in the anterior mandibular corpus of Macaca fascicularis: Implications for models of bone adaptation.

Elastic modulus of bone from the anterior mandibular corpus was determined via microindentation in a mixed-sex ontogenetic sample (N = 14) of Macaca fascicularis. This investigation focused on the hypothesis that material heterogeneity in the macaque mandibular symphysis-provided an accounting of age and sex variation-is explicable as a means to homogenize strains in this region. Experimental data and theoretical models of masticatory loading indicate that in the absence of material compensation, large strain gradients exist in the anterior mandibular corpus of macaques, particularly between lingual and labial cortical plates owing to the effects of lateral transverse bending. Microindentation data indicate that juvenile macaques possess less stiff bone than their subadult and adult counterparts; however, sex differences in elastic modulus are not apparent. Anisotropy variation is idiosyncratic; that is, there is not a common pattern of variation in stiffness sampled among orthogonal planes across individuals. Similarly, differences in stiffness between lingual and labial cortical plates, as well as differences among alveolar, midcorpus, and basal regions are inconsistently observed. Consequently, we find little evidence in support of the hypothesis that spatial variation in bone stiffness functions to homogenize strains in the anterior corpus; in fact, in some individuals, this spatial variation serves to exacerbate, rather than to minimize, strain gradients. The mechanical benefit of elastic modulus variation in the macaque mandibular symphysis is unclear; this variation may not confer adaptive benefit in terms of structural integrity despite the fact that such variation has discernible functional consequences.

Geometric properties and comparative biomechanics of Homo floresiensis mandibles.

The hypodigm of Homo floresiensis from the cave of Liang Bua on Flores Island in the archipelago of Indonesia includes two mandibles (LB1/2 and LB6/1). The morphology of their symphyses and corpora has been described as sharing similarities with both australopiths and early Homo despite their Late Pleistocene age. Although detailed morphological comparisons of these mandibles with those of modern and fossil hominin taxa have been made, a functional analysis in the context of masticatory biomechanics has yet to be performed. Utilizing data on cortical bone geometry from computed tomography scans, we compare the mechanical attributes of the LB1 and LB6 mandibles with samples of modern Homo, Pan, Pongo, and Gorilla, as well as fossil samples of Paranthropus robustus, Australopithecus africanus and South African early Homo. Structural stiffness measures were derived from the geometric data to provide relative measures of mandibular corpus strength under hypothesized masticatory loading regimes. These mechanical variables were evaluated relative to bone area, mandibular length and estimates of body size to assess their functional affinities and to test the hypothesis that the Liang Bua mandibles can be described as scaled-down variants of either early hominins or modern humans. Relative to modern hominoids, the H. floresiensis material appears to be relatively strong in terms of rigidity in torsion and transverse bending, but is relatively weak under parasagittal bending. Thus, they are 'robust' relative to modern humans (and comparable with australopiths) under some loads but not others. Neither LB1 nor LB6 can be described simply as 'miniaturized' versions of modern human jaws since mandible length is more or less equivalent in Homo sapiens and H. floresiensis. The mechanical attributes of the Liang Bua mandibles are consistent with previous inferences that masticatory loads were reduced relative to australopiths but remained elevated relative to modern Homo.

Ontogeny of material stiffness heterogeneity in the macaque mandibular corpus.

Evidence is accumulating that bone material stiffness increases during ontogeny, and the role of elastic modulus in conditioning attributes of strength and toughness is therefore a focus of ongoing investigation. Developmental changes in structural properties of the primate mandible have been documented, but comparatively little is known about changes in material heterogeneity and their impact on biomechanical behavior. We examine a cross-sectional sample of Macaca fascicularis (N = 14) to investigate a series of hypotheses that collectively evaluate whether the patterning of material stiffness (elastic modulus) heterogeneity in the mandible differs among juvenile, subadult and adult individuals. Because differences in age-related activity patterns are known to influence bone stiffness and strength, these data are potentially useful for understanding the relationship between feeding behavior on the one hand and material and structural properties of the mandible on the other. Elastic modulus is shown to be spatially dependent regardless of age, with this dependence being explicable primarily by differences in alveolar versus basal cortical bone. Elastic modulus does not differ consistently between buccal and lingual cortical plates, despite likely differences in the biomechanical milieu of these regions. Since we found only weak support for the hypothesis that the spatial patterning of heterogeneity becomes more predictable with age, accumulated load history may not account for regional differences in bone material properties in mature individuals with respect to the mandibular corpus.

Dietary variation and food hardness in sooty mangabeys (Cercocebus atys): implications for fallback foods and dental adaptation.

We present information on food hardness and monthly dietary changes in female sooty mangabeys (Cercocebus atys) in Tai Forest, Ivory Coast to reassess the hypothesis that thick molar enamel is parsimoniously interpreted as a response to consumption of hard foods during fallback periods. We demonstrate that the diet of sooty mangabeys varies seasonally, but that one food--Sacoglottis gabonensis--is the most frequently consumed food every month and year round. This food is the hardest item in the sooty diet. Given that this species has among the thickest enamel within the primate order, a plausible conclusion is that thick enamel in this taxon evolved not in response to seasonally critical function or fallback foods, but rather to the habitual, year round processing of a mechanically protected foodstuff. These data serve as a caution against de rigueur interpretations that reliance on fallback foods during lean periods primarily explains the evolution of thick enamel in primates.

The relationship between bending stress and the shape of maxillary canines in cercopithecoid monkeys.

Primate canines function in social displays but are also recruited for biting in agonistic encounters. Although the precise nature of the loads such behavior places on the canine crown is unknown, it is thought that bending is a major component of such loads. To date, modeling of canine bending strength has relied on idealized geometric representations. Accounting for the tapering of the crown as well as shape changes along an apical-basal axis provides a more realistic model for assessment of bending stress in canines. We provide such an accounting and evaluate the hypothesis that the morphology of the cercopithecoid canine represents a structural solution for maintaining constant maximum bending stress under apical or distributed loading in a parasagittal plane. This isostress hypothesis is analogous to a design criterion of minimum mass for a given structural requirement. Examining permanent maxillary canines from males and females representing eight West African cercopithecoid species, we reconstructed crown geometry from apex to base using microcomputed tomography. From reconstructed cross-sections, we determined section moduli about a buccolingual centroidal axis. We then determined what the taper should be for a variety of parasagittal loading distributions if the isostress hypothesis were true and compared these theoretical tapers to actual crown geometry. We found that a variety of loading distributions can be accommodated by the canines, particularly among males. These results suggest that our sample of canines are not optimized for resisting particular biting loads, but effectively limit stress gradients associated with a range of behaviors.

Homoplasy and thick enamel in primates.

Traditionally, thick enamel has often been used to infer durophagy (i.e., hard nut and seed consumption) in extinct hominins. These inferences are based on the hypothesis that thick enamel is primarily an adaptation to prevent tooth fracture or chipping resulting from high-stress loads produced during the mastication of large hard foods. An alternative view argues that thick enamel may aid in maintaining tooth function in the face of gradual dental wear from grit, phytoliths and acid, which may be found in foods of widely varying hardness. We use estimates of primate dietary abrasiveness and recorded lifespan to test the hypothesis that enamel thickness is selectively responsive to lifetime dental wear resistance. We use data from the literature to relate enamel thickness to measures of dietary abrasiveness, diet profiles, and longevity for 17 primate species and performed linear regression using several combinations of these variables. We found a positive association between lifetime dietary wear and enamel thickness, suggesting that thick molar enamel in primates may have evolved as a means to resist wear apart from selection to resist tooth fracture. Assuming our estimates of lifetime dietary wear are accurate, we caution against ascribing thick enamel solely to the presence of hard-object feeding in paleoanthropological contexts without also considering primate lifespan and other aspects of feeding ecology.

Dental wear among cercopithecid monkeys of the Taï forest, Côte d'Ivoire.

Studies of dental macrowear can be useful for understanding masticatory and ingestive behavior, life history, and for inferring dietary information from the skeletal material of extinct and extant primates. Such studies to date have tended to focus on one or two teeth, potentially missing information that can be garnered through examination of wear patterns across the tooth row. Our study measured macrowear in the postcanine teeth of three sympatric cercopithecid species from the Taï Forest, Côte d'Ivoire (Cercocebus atys, Procolobus badius, and Colobus polykomos), whose diets have been well-described. Inter-specific analyses suggest that different diets and ingestive behaviors are characterized by different patterns of wear across the molar row, with Cercocebus atys emphasizing tooth use near P4 -M1 , P. badius emphasizing a large amount of tooth use near M2 -M3 , and Colobus polykomos exhibiting wear more evenly across the postcanine teeth. Information regarding differential tooth use across the molar row may be more informative than macrowear analysis of isolated teeth for making inferences about primate feeding behavior.

Viewpoints: feeding mechanics, diet, and dietary adaptations in early hominins.

Inference of feeding adaptation in extinct species is challenging, and reconstructions of the paleobiology of our ancestors have utilized an array of analytical approaches. Comparative anatomy and finite element analysis assist in bracketing the range of capabilities in taxa, while microwear and isotopic analyses give glimpses of individual behavior in the past. These myriad approaches have limitations, but each contributes incrementally toward the recognition of adaptation in the hominin fossil record. Microwear and stable isotope analysis together suggest that australopiths are not united by a single, increasingly specialized dietary adaptation. Their traditional (i.e., morphological) characterization as "nutcrackers" may only apply to a single taxon, Paranthropus robustus. These inferences can be rejected if interpretation of microwear and isotopic data can be shown to be misguided or altogether erroneous. Alternatively, if these sources of inference are valid, it merely indicates that there are phylogenetic and developmental constraints on morphology. Inherently, finite element analysis is limited in its ability to identify adaptation in paleobiological contexts. Its application to the hominin fossil record to date demonstrates only that under similar loading conditions, the form of the stress field in the australopith facial skeleton differs from that in living primates. This observation, by itself, does not reveal feeding adaptation. Ontogenetic studies indicate that functional and evolutionary adaptation need not be conceptually isolated phenomena. Such a perspective helps to inject consideration of mechanobiological principles of bone formation into paleontological inferences. Finite element analysis must employ such principles to become an effective research tool in this context.