Matthew W. Tocheri

How and why hominins evolved key behaviors, such as obligate bipedalism and sophisticated tool production, are key questions in paleoanthropology. Crucial to this inquiry are accurate reconstructions of the morphology and behavior that characterized the panin–hominin last common ancestor (LCA). Because all extant hominines besides Homo sapiens knuckle‐walk, this rare hand posture has received special attention as a possible behavior in the LCA's locomotor repertoire. Arguments concerning knuckle‐walking ancestry in hominins typically examine functional morphology to explore possible adaptations to knuckle‐walking or leverage biological principles to evaluate the likelihood of certain evolutionary scenarios. Because wrist biomechanics in knuckle walking differ from those of other behaviors, carpals are frequently investigated for possible knuckle‐walking specializations. Here, we quantified carpal morphology using spherical harmonics analysis to reevaluate purported knuckle‐walking adaptations among an anthropoid sample with variable locomotor behaviors. We calculated Euclidean distances between behavioral pairs to investigate whether the carpals in knuckle‐walking taxa appear specialized relative to those of other taxa. Results show that nearly all hypothesized knuckle‐walking adaptations are broadly present in anthropoids, and several are specially shared with Asian apes. Some African ape carpal features are shared uniquely with humans and may signal knuckle‐walking ancestry. These results overall suggest that (1) the carpals of African apes are not specialized for knuckle walking; (2) to evolve knuckle walking, only few morphological changes to ancestral hominid carpals are necessary; and (3) an African ape‐like carpus is not too specialized to be a viable ancestral form from which the hominin carpus evolved.

Objectives: Although there has been copious research on the postcranial functional morphology of Homo floresiensis , there has been relatively little study of its pelvis and its implications for interpretations of H. floresiensis locomotion. Here we present the first comparative morphometric analysis of the pelvis of LB 1, the holotype of H. floresiensis , using data derived from high‐resolution micro‐CT scans and situated within a broad, Plio‐Pleistocene hominin context. This includes a novel quantitative measure of iliac flare in hominins to assess the australopith‐like qualities of the LB 1 pelvis described by previous studies. Materials and Methods: In a sample of extant humans and African apes and 26 fossil hominin pelves, we quantified linear and angular pelvic traits, with a focus on features that should provide functional insights into H. floresiensis locomotor behavior, including ilium orientation angle and lower ilium length. Morphometric and allometric hypotheses were investigated using Kruskal–Wallis tests, principal component analysis, and least squares regression. Results: Univariate analyses demonstrate that all pelvic features of LB 1 (except for ischium length) are within the ranges of variation exhibited by Homo sapiens . Multivariate analysis reveals that the H. floresiensis pelvis shares similarities with those of other Pleistocene Homo and some australopith individuals. Discussion: Overall, our results suggest that the pelvis of LB 1 is more Homo ‐like than previously described. While the mix of primitive and derived traits in the postcrania of H. floresiensis makes functional inferences challenging, our results from the pelvis suggest a human‐like form of terrestrial bipedalism in this species.

Since its discovery, Homo floresiensis —an extinct, short-statured, and small-brained hominin species from Flores, Indonesia—has often been ascribed unexpectedly advanced behaviors, such as hunting large game and using fire. Here, we report the results of a systematic taphonomic study sampling the proboscidean bone assemblage at Liang Bua where the frequency and locations of predatory marks, along with skeletal part profiles, show that Komodo dragons likely had primary access to these remains leaving behind only low-utility elements for H. floresiensis to scavenge. Moreover, no signs of intentional use of fire are present in the stratigraphic units associated with H. floresiensis . Together, these results suggest that H. floresiensis was not as behaviorally advanced as originally suggested and provides critical insights into the behavioral repertoire of H. floresiensis , raising important questions about its ancestry.

Primates adopt a variety of hand postures during an impressive diversity of locomotor and manipulative behaviors. Morphological research has found that elements of the hand skeleton, such as the hamate, hold key information for inferring hand use and locomotor kinematics in extinct species. However, debate remains surrounding how and when specialized locomotor hand postures evolved. This debate persists due to uncertainty concerning which taxa possess derived character states for given traits and whether those states evolved via selective response to functional demands or are the result of other biological processes, such as developmental constraint or phylogeny. Phylogenetic comparative methods provide one way to disentangle the role of phylogeny in shaping skeletal morphology and thus can provide clarity concerning which traits arose via other means, such as directed selection. Here, we quantify several functionally relevant hamate traits and employ phylogenetic methods (bayou, OUwie) to infer if and how selection has shaped these traits in particular taxa or clades. We find that selection caused the hamulus to reduce in tarsiers and to lengthen in cebids and hominoids, potentially related to increased demand for grip capability during behaviors like tool use or—in hominoids—below‐branch suspension. Evidence from traits associated with suspension suggests that this behavior evolved independently in hylobatids and hominids. Shifts toward shorter hamate body lengths in Gorilla and hominins cannot be explained by phylogeny alone. Variability among hominines eschews simple interpretation, but this result suggests that if knuckle‐walking evolved once in hominines, then the hamate subsequently lengthened in Pan .

Drimolen Main Quarry (the primary excavation within the Drimolen paleocave system in Gauteng Province, South Africa) dates to ~2 Ma and has yielded abundant hominin remains. Most craniodental specimens are attributed to Paranthropus robustus , but the collection also includes a cranial vault of Homo aff. erectus and teeth with early Homo affinities. Here we catalog isolated postcranial specimens recovered between 1994 and 2015 and contribute formal descriptions along with comparative and functional analyses to provide further information on the paleobiology of the Drimolen hominins. Twenty‐eight fossil specimens are cataloged, although two previously considered as possibly hominin are formally removed from that status. Like the craniodental assemblage, the postcrania include a high proportion of juveniles and subadults. Upper limb elements include an adult distal humerus (DNH 32), ulnar fragment (DNH 109, previously described), three fetal/perinatal humeri, three fetal/perinatal radius fragments, and nine manual phalanges. Lower limb elements include two fetal/perinatal tibiae, hallucal metatarsal, and nonhallucal pedal proximal phalanx. Vertebral and pelvic material include an adult lumbar centrum, a partial pelvis (DNH 43, previously described), and infant thoracic and lumbar neural arches (DNH 150 a and b). DNH 32 and DNH 43 can be attributed to P. robustus with some confidence while DNH 150a and b can possibly be associated with the DNH 134 juvenile cranium attributed to Homo aff. erectus . Although most other specimens probably belong to P. robustus , taxonomic attribution is challenging given Drimolen's mixed assemblage. Functional analyses indicate small‐bodied hominins committed to terrestrial bipedality. However, the first metatarsal evinces a tight tarsometatarsal joint curvature possibly suggesting increased hallucal divergence in that individual. These results might reflect taxonomic variation or evolutionary mosaicism of the postcrania.

Objectives: Phalangeal curvature in hominoids correlates with locomotor behavior, with greater curvature associated with arboreality. Prior research using 2D geometric morphometrics (2DGM) demonstrated this relationship in the third manual proximal phalanges (mPP3) of orangutans. Specifically, the more terrestrial Bornean orangutans ( Pongo pygmaeus ) have less curved mPP3s than the less terrestrial Sumatran orangutans ( Pongo abelii ). This study extends the analysis to the third pedal proximal phalanges (pPP3) to assess if a similar pattern exists and further evaluate the link between curvature and locomotor diversity. Methods: We analyzed pPP3 curvature in P. abelii ( n = 16) and P. pygmaeus ( n = 15) using a 2DGM approach. Comparisons were made with previously published mPP3 data. Generalized Procrustes analyses and principal components analysis were performed on 100 landmarks placed along the diaphysis's central longitudinal axis to quantify shape variation. ANOVA and post hoc tests determined significance between species and autopods. Results: PC1 accounted for 73.4% of shape variance, primarily reflecting dorsoplantar curvature. P. abelii exhibited significantly greater curvature, aligning with its more arboreal lifestyle. Additionally, curvature in pPP3s was greater than in mPP3s within species. Conclusions: Results support the hypothesis that phalangeal curvature reflects habitual substrate use at lower taxonomic resolutions. The greater curvature in P. abelii corresponds to its predominantly arboreal locomotion, while the lesser curvature in P. pygmaeus aligns with increased terrestriality. Additionally, greater curvature in pedal versus manual phalanges may suggest differential adaptations related to substrate engagement. These findings enhance understanding of extant primate autopod functional morphology and can inform reconstructions of extinct species' positional behaviors.

The origin of the human lineage was catalyzed by bipedalism, but how this locomotor mode evolved is debated. We investigated the evolutionary context of human bipedalism by analyzing the morphology of the 4.4 million-year-old hominin talus attributed to Ardipithecus ramidus (ARA-VP-6/500-023). Our results demonstrate that ARA-VP-6/500-023 bears similarities to the tali of chimpanzees and gorillas, who are adapted to vertical climbing and terrestrial plantigrade quadrupedalism. Additionally, we identify the presence of derived features in ARA-VP-6/500-023 consistent with previous suggestions of an enhanced push-off mechanism in the foot of Ar. ramidus. Our observations of the human and ape fossil record are inconsistent with recently proposed models of human origins, which envision the last common ancestor of humans and chimpanzees as a generalized arboreal ape. Instead, our results strongly imply that humans evolved from an African ape-like ancestor, which directly narrows the range of explanations for the origin of our lineage.

When Mary Leakey discovered the OH 5 cranium of Paranthropus boisei alongside Oldowan stone artefacts, it was declared “the oldest yet discovered maker of stone tools”¹. Whether Paranthropus made and used tools has been debated ever since^2,3,4, largely because there are no known hand bones that can be definitively attributed to this genus. Here we report fossil hand and foot bones unambiguously associated with craniodental material of P. boisei. KNM-ER 101000 demonstrates that P. boisei shared key manipulative and bipedal adaptations with the genus Homo. Moreover, the hand morphology of KNM-ER 101000 converges on that of gorillas in ways that are consistent with manual food processing and would have facilitated powerful grasping, such as that used in climbing. These fossils suggest that P. boisei was capable of tool making and use in some capacity while also supporting the proposed dichotomy of distinct dietary adaptations between Paranthropus and Homo. In addition to offering insights into the poorly known postcranial functional anatomy of Paranthropus, this discovery illuminates broader patterns of hominin hand evolution and tool use.

The hindlimb skeleton of the holotype (LB1) of Homo floresiensis is relatively complete and includes both fibulae, which despite being well preserved have yet to be subject to a quantitative comparative analysis with other hominids. A new distal fragment of a fibula has also been recovered from the H. floresiensis‐bearing sediments at Liang Bua (Flores, Indonesia). In this study, we used 3D geometric morphometrics (3DGM) to quantify detailed aspects of the external shape and articular facet morphology of the proximal and distal ends of these H. floresiensis fibulae. The comparative sample included fibulae from 57 extant great apes (Pongo, Gorilla, and Pan), 41 recent and fossil Homo sapiens, five Australopithecus afarensis, and five Neandertals. Shape variation was analyzed using principal component analysis of Procrustes coordinates, and mean differences among taxa were tested using a Procrustes ANOVA with a randomization procedure. Size comparisons were made using centroid size and tested via correlations with principal component scores. Results demonstrate that H. floresiensis fibulae possess the absolute smallest mean linear dimensions and mean centroid sizes among our comparative sample. The proximal and distal fibular ends of H. floresiensis exhibit four key features adapted for obligate bipedalism while also displaying a suite of plesiomorphic traits shared with extant great apes and A. afarensis that, compared with that of H. sapiens and Neandertals, suggest a more versatile ankle joint with a greater range of motion and enhanced load‐bearing capabilities of the fibula. Our results are consistent with other aspects of the H. floresiensis lower limb, such as long feet relative to the femur and a long forefoot relative to the hindfoot, that together suggest an australopith‐like locomotor repertoire that included both obligate bipedalism and climbing.