The Meaning of Neandertal Skeletal Morphology

Home , Mezmaiskaya cave, Neandertal (valley), Shanidar Cave

The meaning of Neandertal skeletal morphology

Timothy D. Weaver1 Department of Anthropology, University of California, One Shields Avenue, Davis, CA 95616; and Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, D-04103 Leipzig, Germany

Edited by Richard G. Klein, Stanford University, Stanford, CA, and approved June 26, 2009 (received for review April 9, 2009)

A procedure is outlined for distinguishing among competing hypotheses for fossil morphology and then used to evaluate current views on the meaning of Neandertal skeletal morphology. Three explanations have dominated debates about the meaning of Nean- dertal cranial features: climatic adaptation, anterior dental loading, and genetic drift. Neither climatic adaptation nor anterior dental loading are well supported, but genetic drift is consistent with the available evidence. Climatic adaptation and activity patterns are the most discussed explanations for Neandertal postcranial features. Robust empirical relationships between climate and body form in extant humans and other endotherms currently make climatic adaptation the most plausible explanation for the wide bodies and relatively short limbs of Neandertals, and many additional postcranial features are likely secondary consequences of these overall skeletal proportions. Activity patterns may explain certain Neandertal postcranial features, but unlike the situation for climate, relationships in extant humans between morphology and activities are typically not well established. For both the cranium and the postcranium, changes in diet or activity patterns may underlie why Neandertals and Pleistocene modern humans tend to be more robust than Holocene humans.

cranium ͉ human evolution ͉ modern human origins ͉ postcranium

ince the discovery of Neandertals ent approaches. I begin by outlining a Lifetime Behavior vs. Evolution. Traits pro- in 1856 at the Kleine Feldhofer procedure for distinguishing among duced by lifetime behaviors should typi- Grotte in the Neander Valley competing hypotheses for fossil mor- cally be found on the skeletons of S near Du¨sseldorf, Germany (1), phology. adults, possibly older subadults, but not deciphering the meaning of Neandertal the skeletons of very young individuals. skeletal morphology has fascinated sci- Distinguishing Among Competing Individuals so young that they have not entists and the public alike. According Explanations yet done certain behaviors would not be to some early proposals, Neandertal The morphology of fossil skeletons, like expected to show traits that result from skeletons were simply pathological mod- any aspect of the phenotype, is the the mechanical loading produced by ac- ern human skeletons, but as more fossils product of genetic influences, environ- tually doing a behavior. This argument accumulated with a consistent set of mental influences, and often interac- is certainly relevant to the foraging, ma- morphological features these explanations between the two (14). Genetic in- nipulative, or mobility activities that tions became untenable (2, 3). Although fluences on skeletal morphology are have been proposed to explain certain pathology cannot explain Neandertal ultimately the result of the adaptive Neandertal traits (16, 17). Therefore, if skeletal morphology in general, patho- (i.e., natural selection) or neutral (i.e., a trait is present on the skeletons of logical lesions, particularly healed trau- mutation, gene flow, genetic drift) evo- young Neandertals, it must either have matic injuries, are frequent on Neander- lutionary forces that have shaped allele an evolutionary explanation (i.e., be due tal skeletons (4). frequencies in a population or species to genetic differences shaped by natural Current evidence suggests that Nean- over multiple generations. There are selection or neutral evolutionary forces) dertals last shared a common ancestor many possible environmental influences or result from environmental influences with modern humans Ͼ350,000 years on the phenotype (14), but dietary, lo- unrelated to adult behaviors. An im- ago (5–7), and fossils that are certainly comotor, or manipulative behaviors that proved understanding of development classified as Neandertals are present in shape skeletal form through the me- often cannot help with distinguishing the fossil record by Ϸ130,000 years ago chanical loading patterns they produce among possible evolutionary explana- (8, 9) and persist until Ͻ35,000 years over the lifetime of an individual are of tions, because neutral evolutionary ago (10, 11). Separate Neandertal and particular interest to investigations of forces or different natural selective pres- modern human lineages perhaps ancient skeletons (15). Therefore, the sures can act through similar develop- emerged when geographic barriers pro- task of deciphering the meaning of Ne- mental shifts (18), but it can help with duced by climate fluctuations isolated andertal skeletal morphology can be deciding whether an evolutionary expla- Neandertal populations in Europe from encapsulated as finding ways to infer the nation is warranted or not. modern human populations further evolutionary and lifetime behavioral Controlled experiments on laboratory south (8, 12). Although Neandertals causes of Neandertal skeletal features. animals are another way to distinguish originated in Europe they later extended This task is not straightforward, and, between lifetime behavioral and evolu- their geographic range into western and even when it is possible to determine tionary explanations for morphology central Asia, ranging as far south as Is- whether genetic or environmental influ- (19, 20). Because it is possible in the rael and perhaps as far east as southern ences are primarily responsible for varia- laboratory to control many variables Siberia (8, 13). tion among individuals, populations, or other than the ones of interest, the re- My purpose is to present current species for a particular skeletal trait, it sults of laboratory experiments can of- views on the meaning of Neandertal is still necessary to decide among com- skeletal morphology. Instead of a giving peting evolutionary or lifetime behav- comprehensive review, I concentrate on ioral explanations. With this in mind, I Author contributions: T.D.W. wrote the paper. aspects of Neandertal cranial and post- outline a procedure for going from a The author declares no conﬂict of interest. cranial morphology that have been stud- description of fossil morphology to its This article is a PNAS Direct Submission. ied extensively using a variety of differ- meaning. 1E-mail: [email protected].

16028–16033 ͉ PNAS ͉ September 22, 2009 ͉ vol. 106 ͉ no. 38 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903864106 Downloaded by guest on September 26, 2021 PCA ETR:PERSPECTIVE FEATURE: SPECIAL

Evolutionary Explanations. If an evolutionary explanation is warranted, the final step is to distinguish among possible hypotheses. One approach is to investigate whether there is an empirical relationship between the morphological feature of interest and a potential selective factor (e.g., temperature, humidity, loco- motor behavior, etc.) in one or more extant species. If so, then perhaps the same relationship explains patterns of variation for the feature in the fossil record. This approach implicitly assumes that whatever factors lead to a relationship in the extant species are also important for the extinct species, and consequently, is most robust when similar relationships are found for multiple populations or species. A second approach is to evaluate whether the form of the feature, or complex of features, observed in fossil specimens is consistent with its pur- Fig. 1. Neandertal and modern human cranial differences. On the left is a Neandertal (La Chapelle- aux-Saints). On the right is a modern human (Cro-Magnon 1). Anterior (above) and lateral (below) views. ported function. This approach does not Photos courtesy of Chris Stringer and the Muse´e de l’Homme (Paris). attempt to directly model how natural selection would have acted; it simply evaluates the internal consistency of an ten be interpreted unambiguously. How- simply due to shared ancestry, even if the adaptive hypothesis. An internally con- ever, it is important to be cautious when features are genetically, developmentally, sistent adaptive hypothesis is not neces- extending laboratory results outside the and functionally unlinked. sarily correct, but an inconsistent one realm of the experiment. It is well An alternative to directly estimating can be rejected. Depending on the par- known that the sources of variation covariance patterns from fossils is to ticular hypothesis, different approaches within and between groups and for dif- study extant species, with the implicit can be taken to assess consistency, but ferent sets of groups can be quite differ- assumption that they have similar co- biomechanical modeling or laboratory experiments are often used (31). ent (21). This means, for example, that variance patterns to the fossil species of Third, it is sometimes possible to dis- variation in a feature could be mostly interest. Covariance patterns are similar tinguish among different evolutionary due to environmental effects for the across human populations (22), and al- laboratory animals within the context of explanations by explicitly modeling evo- though there are some important differ- lutionary forces with quantitative and the experiment, but genetic effects ences, African apes and humans have could be responsible for variation in the population genetics (18, 32). This mod- broadly similar covariance patterns (23– eling can be used to evaluate the impor- feature between different groups of the 25). The sum suggests that it is reason- same species or between different spe- tance of genetic drift vs. natural selec- able to assume that Neandertals would tion or the fit of different selective cies. The central issue is that, although have had similar covariance patterns to hypotheses with observed data. A interpretation of the experimental re- humans, at least for most features. Most strength this approach is that the dy- sults is straightforward, determining important for evolutionary explanations namics of the evolutionary process are their applicability to understanding ex- is the within-group additive genetic co- quantitatively incorporated into the test- tinct species, such as Neandertals, is variance matrix, which can often be ap- ing of hypotheses (33). This additional more difficult. proximated by the within-group pheno- level of quantification requires hypothe- typic covariance matrix (26, 27). Within- ses to be specified more precisely, which Single Features vs. Complexes of Features. group additive genetic covariance is can sometimes be difficult, but, in prin- If features vary together because of ge- ciple, it allows for more rigorous tests. netic, developmental, or functional links, what constrains the response to natural selection or the direction of change by Finally, the patterning of traits in the then explaining one of them may be suffi- fossil record can be used to evaluate cient to explain them all. Ideally, we genetic drift (28). Within-individual covariance patterns (fluctuating asymme- competing evolutionary explanations. would like to be able to estimate the co- The value of this final approach and the try) can give insights into the covariance variance among features directly from fos- previous one would increase substan- caused by developmental interactions sil specimens, but a much larger and less tially if we knew more about how evolu- fragmentary Neandertal sample would be (29). Analyses that combine among- and tionary forces typically act over hun- necessary to robustly estimate covariance within-group variation within a species dreds to thousands of generations. patterns. Even so, arguments that 2 fea- may identify covariance patterns that tures form a complex would be weakened are not readily apparent from analyses Neandertal Cranial Morphology by finding fossil specimens that have one of individual groups (30), but these as- Numerous metric and nonmetric fea- of the features but not the other. How- sociations should be interpreted cau- tures typically distinguish the crania and ever, consistently finding a set of features tiously, because they could be due to mandibles of Neandertals from those of together does not demonstrate that they population history or phylogeny rather modern humans (30, 34–38) (Fig. 1, Ta- all have the same explanation, because than genetic, developmental, or func- ble 1). Table 1 lists both uniquely de- individuals of a species will share features tional links. rived (autapomorphic) and primitive

Weaver PNAS ͉ September 22, 2009 ͉ vol. 106 ͉ no. 38 ͉ 16029 Downloaded by guest on September 26, 2021 Table 1. Selected Neandertal cranial features raised on cooked food showed Ϸ10% less Anatomical region Features growth for some facial dimensions, suggesting that diet, and perhaps other be- Cranial vault Receding frontal squama haviors, may influence facial size (20). Long, low braincase, sometimes with a posteriorly bulging occipital These results are intriguing, but unless the (ЉbunЉ) animals in the 2 treatment groups were Globular (Љen-bombeЉ) braincase when viewed from behind specifically selected to be siblings, they Occipital torus with a suprainiac fossa above it could reflect genetic differences between Cranial base Fairly unﬂexed ectobasicranium treatment groups rather than diet. Al- Large juxtamastoid eminence and relatively small mastoid process though differences in diet do not appear Tubercle on mastoid process adjacent to the external auditory meatus to explain most cranial differences be- Anteroposteriorly elongated foramen magnum tween Neandertals and contemporaneous modern humans, they could explain why Acoustic cavities Large lateral, small anterior, and small and inferiorly positioned the crania of Pleistocene modern humans posterior semicircular canal tend to be more robust than those of Ho- Facial skeleton Pronounced, double-arched supraorbital torus locene humans (19, 20). Projecting midface Minimally angled zygomatic bone Single Features vs. Complexes of Features. Absence of infraorbital concavity and canine fossa On the one hand, studies of extant spe- Wide, tall nasal aperture cies have consistently demonstrated sub- Wide, projecting nasal bridge stantial integration of cranial features Depressed nasal ﬂoor (31). On the other hand, the fossil Mandible Retromolar space record shows a certain degree of inde- Inferiorly positioned and laterally expanded condyle pendence for the cranial features typi- Asymmetric sigmoid notch cally found in Neandertals. For example, Large coronoid process European fossils Ͼ130,000 years old ex- Horizontal-oval mandibular foramen hibit some but not all Neandertal fea- Large medial pterygoid tubercle tures (8), which is not the expected pat- Absence of mental eminence tern if all Neandertal features were part of an integrated package. This point is illustrated nicely with specimens from (symplesiomorphic) features, because Neandertals, along with studies of older the Sima de los Huesos, Spain. Fossils both need to be explained. The appear- subadults (43), is that most Neandertal from this site have prognathic midfaces ance of derived features in the fossil cranial features are present very early in like Neandertals and some of the associ- record can point to the action of direc- development and, consequently, appear ated morphological features, but they tional natural selection or genetic drift, to warrant evolutionary rather than life- lack cranial base traits such as the mas- and the retention of primitive features time behavioral explanations. It should toid tubercle or a large juxtamastoid can indicate stabilizing natural selection. be pointed out, however, that most stud- eminence coupled with a relatively small ies of subadult Neandertals have tended mastoid process (44). They also lack some cranial vault traits, such as a glob- Lifetime Behavior vs. Evolution. Many dis- to focus on the Neandertal features that tinctive Neandertal cranial features are are present rather than providing a sys- ular shape when viewed from behind, but display others, such as an incipient present on the skeletons of very young tematic assessment of the percentages of suprainiac fossa (44). individuals, suggesting that they are not present vs. absent features, which may the result of mechanical loading pat- give the impression that young Neander- Evolutionary Explanations. Three main terns produced by lifetime behaviors. Of tals look more like adult Neandertals evolutionary explanations have been particular interest are the features than they actually do. proposed for Neandertal cranial mor- present on 2 well-preserved neonatal Although many cranial differences be- phology: adaptation to cold climates, skeletons from Mezmaiskaya, Russian tween Neandertals and modern humans adaptation to anterior dental loading, or Federation and Le Moustier, France. are likely due to different allele frequen- genetic drift. The cold-climate hypothe- The Neandertal features found on the cies at loci underlying cranial form, there sis is based on the observation that the Mezmaiskaya specimen include an over- is some experimental evidence that life- geographic range of Neandertals was all cranial shape similar to that of other time behavior differences, specifically centered quite far north in Europe, and Neandertals, an elongated foramen mag- which foods are eaten, can influence cra- except for brief warm periods, global num, a projecting midface, an inferiorly nial form. For example, Lieberman and climates were substantially cooler when positioned posterior semicircular canal, colleagues (20) performed laboratory ex- Neandertals were evolving than they are and an inferiorly positioned mandibular periments on rock hyraxes to investigate today. Neandertals clearly experienced condyle (39, 40). Those on Le Moustier the effects of food processing on cranial fairly cold temperatures, but this does 2 include the absence of an infraorbital growth and form. The maxillary molars of not require that their cranial features concavity and nasal bones shaped like rock hyraxes are positioned directly be- are adaptations to these climatic condi- those of adult Neandertals (41). The neath or behind the orbits as in humans, tions. Most climatic hypotheses for Ne- slightly older Amud 7 skeleton from which may make them more appropriate andertal cranial form focus on the nasal Israel also shows Neandertal cranial models for human mechanical loading region and how other facial features features, including the absence of a patterns than more prognathic nonhuman may result from adaptations in this re- mental eminence, an oval foramen primates (20). The hyraxes were divided gion (30, 36, 45, 46). magnum, and an enlarged medial ptery- into 2 groups of 4 animals each. One Studies by Franciscus (36) of internal goid tubercle (42). The general impres- group was fed cooked food, whereas the nasal dimensions and Holton and Fran- sion from these skeletons of very young other was fed raw/dry food. The animals ciscus (30) of external ones fail to sup-

16030 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903864106 Weaver Downloaded by guest on September 26, 2021 Table 2. Selected Neandertal postcranial features drift could explain Neandertal cranial Anatomical region Features features. My colleagues and I (18) tested this hypothesis using predictions General Wide bodies with short extremities, particularly the distal limb from quantitative and population ge- segments netics with a sample of 37 cranial mea- Long bones tend to have bowed shafts and thick cortical bone surements collected on 20 Neandertal Postcranial bones tend to be robust with rugose muscle attachments specimens and 2,524 recent humans, and we were unable to reject it with Axial skeleton Horizontal lower cervical spinous processes multiple statistical tests. Subsequently, Robust, rounded rib shafts calibrated on the rate of cranial diver- Upper extremity Long clavicle gence among recent human popula- Wide scapula with narrow glenoid fossa and dorsal axillary sulcus tions, we estimated that Neandertals Humerus with wide olecranon fossa and narrow surrounding and modern humans diverged dorsodistal pillars Ϸ311,000 years ago or Ϸ435,000 years Ulna with high and long olecranon process and anteriorly oriented ago, depending on assumptions about trochlear notch within-group variation (7). These Radius with medially directed tuberosity and long neck split dates match quite closely with Trapezium with ﬂat ﬁrst metacarpal facet those derived from ancient Neandertal Hamate with large hamulus and extant human DNA sequences, Third metacarpal with short styloid process which is the expected result if genetic Subequal proximal and distal thumb phalanges drift were responsible for the cranial Large hand distal phalangeal tubercles divergence. Lower extremity Wide pelvis with long, thin superior pubic ramus Additional support for the genetic drift Femur with large articulations, rounded midshaft lacking a pilaster, hypothesis comes from the fossil record. and low neck-shaft angle Neandertal features do not appear all at Tibia with projecting tuberosity and absence of diaphyseal once. In fact, they seem to gradually accu- concavities mulate over a period of Ͼ300,000 years Relatively symmetrical medial and lateral patellar facets (8). A similar pattern may characterize the Large foot distal phalangeal tubercles appearance of modern human features (55), but the dating and more fragmentary nature of the African fossil record port adaptation to cold climates as the pear early in development, they cannot allows for other interpretations. This ‘‘ac- primary explanation for Neandertal fa- be direct mechanical responses to ante- cretion’’ of Neandertal features is exactly cial morphology. The depressed nasal rior dental loading. They would have to the expected pattern if genetic drift were floors often found in Neandertals are be adaptations produced by natural se- responsible, and it seems less compatible most common in recent humans from lection after the species consistently with existing adaptive hypotheses such as subSaharan Africa (36). Likewise, wide performed this behavior for multiple climatic adaptation or anterior dental nasal apertures, which characterize Ne- generations. loading. andertals, are most frequently found in One problem with the anterior dental Neandertal Postcranial Morphology equatorial recent humans (30). Al- loading hypothesis is that biomechanical though the Neandertal nasal region, in modeling suggests that Neandertals were The postcranial anatomy of Neandertals, like their cranial anatomy, distinguishes general, does not appear to be an adap- not able to produce particularly high them from modern humans (34, 38, 56– tation to cold climates, the narrow supe- bite forces (53, 54). Neandertal cranial 59) (Table 2). As with Table 1 for cranial rior internal nasal dimensions, tall nasal form cannot be adapted to resisting high features, Table 2 includes both primitive apertures, and projecting nasal bridges bite forces if Neandertals were incapa- and derived postcranial features. of Neandertals could be, because these ble of producing them in the first place. features are typically found in high-lati- O’Connor and colleagues (54) showed Lifetime Behavior vs. Evolution. At least tude recent humans (30). If Neandertals that, although Neandertals would have some Neandertal postcranial features were adapting to the cold similarly to been able to produce fairly high bite are present on Neandertal fossils from present-day humans, then climatic adap- forces in absolute magnitude, their bite individuals Ͻ1 year of age. For example, tation is an unlikely explanation for forces would not have been unusually Mezmaiskaya has a femoral diaphysis their cranial form. large for the size of their crania. Addi- that is long relative to the tibial diaphy- Neandertals tend to have more worn tionally, if efficiency is quantified as the sis, bowed long bones, a long superior anterior teeth than posterior ones, and ratio of bite force to muscle force, Ne- pubic ramus of the pelvis, a medially their anterior teeth show a high inci- andertals were actually less efficient directed radial tuberosity, and a fairly dence of enamel chipping, microfrac- than many modern humans (54). robust skeleton (39, 40). Additionally, tures, and microstriations on the labial Neandertal and modern human pop- Amud 7 has a long clavicle (60), and the surfaces. Taken together, these signa- ulations seem to have became isolated Kiik-Koba 2 individual from the Crimea, tures of anterior tooth use suggest that from each other Ͼ350,000 years ago thought to be Ϸ3–7 months old, has an Neandertals were using their mouths (5–7). When this occurred, Neandertals incipient scapular dorsal axillary sulcus like a vise. The anterior dental loading and modern humans would have di- (61) and an opponens pollicis flange on hypothesis extends this idea by propos- verged from each other even in the the first metacarpal (56). However, ing that Neandertal facial form, and absence of natural selection through some Neandertal postcranial features perhaps other cranial features, are adap- the random fluctuations in allele fre- only appear later in development, in- tations to dissipate the high mechanical quencies that happen in all real popu- cluding the thinness of the superior pu- loads produced by this behavior (47–52). lations (i.e., populations that are not bic ramus (62) and thick long-bone cor- Because Neandertal facial features ap- infinite in size). This process of genetic tices (60, 63).

Weaver PNAS ͉ September 22, 2009 ͉ vol. 106 ͉ no. 38 ͉ 16031 Downloaded by guest on September 26, 2021 Numerous studies of laboratory ani- out, the percentage of the remaining vari- Even if a wide pelvis was unrelated to mals, human growth series, and biome- ance explained by the integration model climate in early Homo, climate adaptation chanical models indicate that long bone might be quite high. is still the best explanation for why Nean- cortical thickness reflects, at least in In a study of the entire postcranial skele- dertals maintained a wide pelvis, early part, mechanical loading produced by ton, Pearson (69) found that articular size modern humans living closer to the equa- locomotion and other behaviors (15, and long bone shaft thickness relative to tor evolved a narrow pelvis, and recent 64). These results are consistent across a bone length is closely related to body pro- humans who migrated to cold climates variety of studies, making it likely that portions. Focusing more narrowly on the hip regained a wide pelvis. mechanical loading history explains a region, I (71) found that a wide pelvis rela- Although climatic adaptation may ex- significant portion of the variation in tive to femur length was associated with plain Neandertal body proportions and cortical bone thickness between Nean- femora with large articulations, thick and other morphological features of their post- dertals and modern humans, but it is round midshafts shafts, and low neck-shaft cranium, activity patterns could be re- possible that at least some differences angles. Neandertals typically have large artic- sponsible for at least some postcranial are due to genetic influences. ulations, thick shafts, and femora with round features, because, as discussed above, nu- midshafts and low neck-shaft angles, so merous studies indicate that long bone Single Features or Complexes of Features. based on the results of these 2 studies, it is cortical thickness reflects mechanical One of the foremost postcranial contrasts plausible that many Neandertal postcranial loading from locomotion and other behav- between Neandertals and early modern features are best explained as secondary iors. As with the cranium, activity levels humans is in body proportions. Specifi- consequences of their body proportions. Fur- may also explain why Pleistocene modern cally, Neandertals have a wide pelvis, thermore, a wide pelvis coupled with the humans tend to have more robust postcra- short limbs relative to trunk height, and maintenance of a transversely oval outlet of nia than Holocene humans (15, 58). short distal limb segments, whereas early the birth canal may explain additional Nean- modern humans have narrower bodies dertal pelvic features, including why both Conclusions with relatively longer limbs (45, 57, 58, sexes have similar superior pubic ramus Deciphering the meaning of Neandertal 65–69). These contrasts in body propor- lengths (72). skeletal morphology is a complex en- tions have lead multiple researchers to deavor that starts with determining propose that other postcranial differences Evolutionary and Lifetime Behavioral Expla- whether a morphological feature is best between Neandertals and modern humans nations. Two main evolutionary explana- considered to result from lifetime behav- are simply secondary consequences of tions have been proposed for Neandertal iors or, alternatively, from alleles passed differences in body proportions. If this postcranial morphology: adaptation to from parents to offspring that were previ- proposal is correct, an explanation for cold climates or activity patterns (either ously shaped by evolutionary forces. It Neandertal body proportions could ac- lifetime behavioral response or evolution- continues with investigations of whether a count for many other postcranial features ary adaptation). In recent humans (65– feature is best explained in isolation or as as well. There are few fossil individuals 67), as in other endothermic species (73, part of a complex of features. It concludes that preserve a large fraction of their 74), body proportions like those of Nean- with evaluating competing hypotheses us- postcranial skeleton, so, unlike the situa- dertals are characteristic of individuals ing empirical studies of extant species, tion with cranial features, covariance pat- with ancestry in cold climates. Individuals tests of form vs. function, modeling based terns among postcranial features are diffi- with ancestry in warm climates, in con- on quantitative and population genetics, cult to investigate directly with the fossil trast, tend to have similar body propor- and documenting the patterning of fea- record. Consequently, I focus on studies tions to the earliest modern humans (68). tures in the fossil record. of covariance patterns based primarily on These robust empirical patterns suggest Currently, the best explanation for recent human samples. that Neandertal body proportions, and the many Neandertal cranial features is diver- Churchill (70) investigated whether up- covarying features discussed in the previ- gence by genetic drift that began when per body features typically found in Nean- ous section, reflect adaptation to the cold Neandertal and modern human popula- dertals could be explained as secondary climates of Pleistocene Eurasia. tions became isolated from each other consequences of upper body size, chest The climate hypothesis is further Ͼ350,000 years ago. This explanation is shape, and robusticity. He found fairly supported by laboratory experiments. supported by modeling based on quantita- weak correlations between upper body For example, Tilkens and colleagues tive and population genetics and the ‘‘ac- variables. Although a model of upper (75) showed that human subjects with cretional’’ appearance of Neandertal fea- body integration fit the data better than long legs relative to body mass had tures in the fossil record. Neandertal body one with no integration, the integration higher resting metabolic rates than in- proportions, and likely other postcranial model explained less than half of the vari- dividuals with relatively shorter limbs features, appear to be adaptations, or sec- ance in upper body morphology. Based on when sitting in an Ϸ22 °C room wear- ondary consequences of adaptations, to these results, Churchill (70) concluded ing shorts and a T-shirt. These experi- the typically cold climates of Pleistocene that it was unlikely that an overarching ments demonstrate that individuals Eurasia. However, for both the cranium causal factor, such as body proportions, with ‘‘warm adapted’’ body proportions and the postcranium, changes in diet or could explain Neandertal upper body need to expend more energy to pre- activity patterns may explain some fea- morphology. This interpretation is reason- vent their body temperatures from tures, and in particular, could underlie able, but it is important to note that only dropping due to heat loss. why Neandertals and Pleistocene modern a fraction of the unexplained variance is It could be argued that the early Homo humans are more robust than Holocene likely to be the result of lifetime behav- pelvis from Gona, Ethiopia refutes the humans. ioral or evolutionary causes. Some propor- climate hypothesis, because it may demon- tion of phenotypic variance will always be strate that a wide pelvis was the primitive ACKNOWLEDGMENTS. I thank Richard Klein and due to genetic variation among individu- condition for the genus Homo (76). How- David Stopak for inviting me to contribute to the als, microenvironmental variation, random ever, showing that a morphological fea- special feature on ‘‘Out of Africa: Modern Human Origins’’; Chris Stringer and the Muse´e de l’Homme developmental perturbations, and mea- ture is primitive for a taxonomic group (Paris) for the photos for Fig. 1; and Jean-Jacques surement error (14). If the variance due does not explain why this feature persists Hublin, Teresa Steele and 2 anonymous reviewers to these factors were to be partitioned in some descendant taxa and not others. for helpful feedback.

16032 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903864106 Weaver Downloaded by guest on September 26, 2021 1. Schaaffhausen H (1857) Brief communication (Trans- 26. Cheverud JM (1988) A comparison of genetic and phe- 52. Trinkaus E (1987) The Neandertal face: Evolutionary lated from German). Verhandlungen des Naturhisto- notypic correlations. Evolution 42:958–968. and functional perspectives on a recent hominid face. J rischen Vereins der Preussischen Rheinlande und West- 27. Roff DA (1996) The evolution of genetic correlations: Hum Evol 16:429–443. falens 14:50–52. An analysis of patterns. Evolution 50:1392–1403. 53. Antoń SC (1994) In Integrative Paths to the Past: Pa- 2. Stringer CB, Gamble C (1993) In Search of the Nean- 28. Lande R (1979) Quantitative genetic analysis of multi- leoanthropological Advances in Honor of F. Clark derthals: Solving the Puzzle of Human Origins (Thames variate evolution, applied to brain:body size allometry. Howell, eds Corruccini RS, Ciochon RL (Prentice Hall, and Hudson, New York). Evolution 33:402–416. Englewood Cliffs, NJ), pp 677–695. 3. Trinkaus E, Shipman P (1992) The Neandertals: Of Skel- 29. Klingenberg CP, Badyaev AV, Sowry SM, Beckwith NJ 54. O’Connor CF, Franciscus RG, Holton NE (2005) Bite force etons, Scientists, and Scandal (Random House, New (2001) Inferring developmental modularity from mor- production capability and efficiency in Neandertals York). phological integration: Analysis of individual variation and modern humans. Am J Phys Anthropol 127:129– 4. Berger T, Trinkaus E (1995) Patterns of trauma among and asymmetry in bumblebee wings. Am Nat 157:11– 151. the Neandertals. J Archaeol Sci 22:841–852. 23. 55. Braüer G (2008) The origin of modern anatomy: By 5. Stringer CB, Hublin J-J (1999) New age estimates for the 30. Holton NE, Franciscus RG (2008) The paradox of a wide speciation or intraspecific evolution? Evol Anthropol Swanscombe hominid, and their significance for hu- nasal aperture in cold-adapted Neandertals: A causal 17:22–37. man evolution. J Hum Evol 37:873–877. assessment. J Hum Evol 55:942–951. 56. Trinkaus E (1983) In The Mousterian Legacy: Human 6. Noonan JP, et al. (2006) Sequencing and analysis of 31. Lieberman DE (2008) Speculations about the selective Biocultural Change in the Upper Pleistocene,ed Neanderthal genomic DNA. Science 314:1113–1118. basis for modern human craniofacial form. Evol An- Trinkaus E (British Archaeological Reports Interna- 7. Weaver TD, Roseman CC, Stringer CB (2008) Close cor- thropol 17:55–68. tional, Oxford), Vol 164, pp 165–200. respondence between quantitative- and molecular- 32. Ackermann RR, Cheverud JM (2004) Detecting genetic 57. Holliday TW (1997) Postcranial evidence of cold adap- genetic divergence times for Neandertals and modern drift versus selection in human evolution. Proc Natl tation in European Neandertals. Am J Phys Anthropol humans. Proc Natl Acad Sci USA 105:4645–4649. Acad Sci USA 101:17946–17951. 104:245–258. 8. Hublin J-J (1998) In Neandertals and Modern Humans 33. Roseman CC, Weaver TD (2007) Molecules versus mor- 58. Pearson OM (2000) Postcranial remains and the origins in Western Asia, eds Akazawa T, Aoki K, Bar-Yosef O phology? Not for the human cranium. Bioessays of modern humans. Evol Anthropol 9:229–247. (Plenum, New York), pp 295–310. 29:1185–1188. 59. Yokley TR, Churchill SE (2006) Archaic and modern 9. Rink WJ, Schwarcz HP, Smith FH, Radovcic J (1995) ESR 34. Franciscus RG (2002) Encyclopedia of Evolution,ed human distal humeral morphology. J Hum Evol 51:603– ages for Krapina hominids. Nature 378:24. Pagel M (Oxford Univ Press, Oxford), pp 493–497. 10. Hublin J-J, Barroso Ruiz C, Medina Lara P, Fontugne M, 616. 35. Rak Y, Ginzburg A, Geffen E (2002) Does Homo nean- 60. Odwak H (2000) Long bone robusticity and claviculo- Reyss J-L (1995) The Mousterian site of Zafarraya (An- derthalensis play a role in modern human ancestry? dalucia, Spain): Dating and implications on the Palaeo- humeral proportions of the Amud 7 Neandertal baby. The mandibular evidence. Am J Phys Anthropol lithic peopling processes of Western Europe. Comptes Am J Phys Anthropol 30 (Suppl):241. 119:199–204. Rendus de l’Academie des Sciences Paris 321:931–937. 61. Trinkaus E (2008) Kiik-Koba and Neandertal axillary 36. Franciscus RG (2003) Internal nasal floor configuration 11. Higham T, Bronk Ramsey C, Karavanic I, Smith FH, border ontogeny. Anthropol Sci 116:231–236. in Homo with special reference to the evolution of Trinkaus E (2006) Revised direct radiocarbon dating of 62. Tompkins RL, Trinkaus E (1987) La Ferrassie 6 and the Neandertal facial form. J Hum Evol 44:701–729. the Vindija G Upper Paleolithic Neandertals. Proc Natl development of Neandertal pubic morphology. Am J 1 37. Spoor F, Hublin J-J, Braun M, Zonneveld F (2003) The Acad Sci USA 103:553–557. Phys Anthropol 73:233–239. bony labyrinth of Neanderthals. J Hum Evol 44:141– 12. Howell FC (1952) Pleistocene glacial ecology and the 63. Cowgill LW, Trinkaus E, Zeder MA (2007) Shanidar 10: 165. evolution of ‘‘classic Neandertal man.’’ Southwestern J A Middle Paleolithic immature distal lower limb from 38. Trinkaus E (2006) Modern human versus Neandertal Anthropol 8:377–410. Shanidar Cave, Iraqi Kurdistan. J Hum Evol 53:213–223. evolutionary distinctiveness. Curr Anthropol 47:597– 13. Krause J, et al. (2007) Neanderthals in central Asia and 64. Carter DR, Beaupre´ GS (2001) Skeletal Function and 620. Siberia. Nature 449:902–904. Form: Mechanobiology of Skeletal Development, Ag- 39. Golovanova LV, Hoffecker JF, Kharitonov VM, Ro- 14. Lynch M, Walsh B (1998) Genetics and Analysis of ing, and Regeneration (Cambridge Univ Press, Cam- manova G (1999) Mezmaiskaya Cave: A Neanderthal Quantitative Traits (Sinauer, Sutherland, MA). bridge, UK). occupation in the northern Caucasus. Curr Anthropol 15. Ruff CB (2000) In Biological Anthropology of the Hu- 65. Trinkaus E (1981) In Aspects of human evolution, Sym- 40:77–86. man Skeleton, eds Katzenberg MA, Saunders SR posia of the society for the study of human biology,ed 40. Ponce de Leoń M, et al. (2008) Neanderthal brain size at (Wiley-Liss, New York), 2nd Ed, pp 71–102. Stringer CB (Taylor and Francis, London), Vol 21, pp birth provides insights into the evolution of human life 16. Trinkaus E (1997) Appendicular robusticity and the pa- 187–224. history. Proc Natl Acad Sci USA 105:13764–13768. leobiology of modern human emergence. Proc Natl 66. Ruff CB (1994) Morphological adaptation to climate in 41. Maureille B (2002) A lost Neanderthal neonate found. Acad Sci USA 94:13367–13373. modern and fossil hominids. Yearbook Phys Anthropol 17. Schmitt D, Churchill SE, Hylander WL (2003) Experimen- Nature 419:33–34. 42. Rak Y, Kimbel WH, Hovers E (1994) A Neandertal infant 37:65–107. tal evidence concerning spear use in Neandertals and 67. Holliday TW (1997) Body proportions in Late Pleisto- early modern humans. J Archaeol Sci 30:103–114. from Amud Cave, Israel. J Hum Evol 26:313–324. 43. Ponce de Leoń MS, Zollikofer CPE (2001) Neanderthal cene Europe and modern human origins. J Hum Evol 18. Weaver TD, Roseman CC, Stringer CB (2007) Were Ne- 32:423–447. andertal and modern human cranial differences pro- cranial ontogeny and its implications for late hominid 68. Holliday TW (2000) Evolution at the crossroads: Mod- duced by natural selection or genetic drift? J Hum Evol diversity. Nature 412:534–538. ern human emergence in Western Asia. Am Anthropol 53:135–145. 44. Arsuaga JL, Martínez I, Gracia A, Lorenzo C (1997) The 102:54–68. 19. Lieberman DE (1996) How and why humans grow thin Sima de los Huesos crania (Sierra de Atapuerca, Spain). 69. Pearson OM (2000) Activity, climate, and postcranial skulls: Experimental evidence for sytemic cortical ro- A comparative study. J Hum Evol 33:219–281. robusticity. Curr Anthropol 41:569–607. busticity. Am J Phys Anthropol 101:217–236. 45. Coon CS (1962) The Origin of Races (Alfred A. Knopf, 70. Churchill SE (1996) Particulate versus integrated evolu- 20. Lieberman DE, Krovitz GE, Yates FW, Devlin M, St. New York). tion of the upper body in Late Pleistocene humans: A Claire M (2004) Effects of food processing on mastica- 46. Sergi S (1958) In Hundert Jahr Neanderthaler,edvon test of two models. Am J Phys Anthropol 100:559–583. tory strain and craniofacial growth in a retrognathic Koenigswald GHR (Kemink en Zoon, Utrecht), pp 38– 71. Weaver TD (2003) The shape of the Neandertal femur face. J Hum Evol 46:655–677. 51. is primarily the consequence of a hyperpolar body 21. Feldman M, Lewontin RC (1975) The heritability hang- 47. Brace CL (1964) The fate of the ‘‘classic’’ Neanderthals: up. Science 190:1163–1168. A consideration of hominid catastrophism. Curr An- form. Proc Natl Acad Sci USA 100(12):6926–6929. 22. Gonza´les-Jose´ R, Van der Molen S, Gonzales-Perez E, thropol 5:3–43. 72. Weaver TD, Hublin J-J (2009) Neandertal birth canal Hernandez M (2004) Patterns of phenotypic covaria- 48. Demes B (1987) Another look at an old face: Biome- shape and the evolution of human childbirth. Proc Natl tion and correlation in modern humans as viewed from chanics of the Neandertal facial skeleton reconsidered. Acad Sci USA 106:8151–8156. morphological integration. Am J Phys Anthropol J Hum Evol 16:297–303. 73. Ashton KG, Tracy MC, de Queiroz A (2000) Is Berg- 123:69–77. 49. Rak Y (1986) The Neanderthal: A new look at an old mann’s rule valid for mammals? Am Nat 156:390–415. 23. Ackermann RR (2005) Ontogenetic integration of the face. J Hum Evol 15:151–164. 74. Freckleton RP, Harvey PH, Pagel M (2003) Bergmann’s hominoid face. J Hum Evol 48:175–197. 50. Smith FH, Paquette SP (1989) In The Emergence of rule and body size in mammals. Am Nat 161:821–825. 24. Polanski JM, Franciscus RG (2006) Patterns of craniofa- Modern Humans: Biocultural Adaptations in the Later 75. Tilkens MJ, Wall-Scheffler CM, Weaver TD, Steudel- cial integration in extant Homo, Pan, and Gorilla. Am J Pleistocene, ed Trinkaus E (Cambridge Univ Press, Cam- Numbers K (2007) The effects of body proportions on Phys Anthropol 131:38–49. bridge, UK), pp 181–210. thermoregulation: An experimental assessment of 25. Ackermann RR (2009) Morphological integration and 51. Spencer MA, Demes B (1993) Biomechanical analysis of Allen’s rule. J Hum Evol 53:286–291. the interpretation of fossil hominin diversity. Evol Biol masticatory system configuration in Neandertals and 76. Simpson SW, et al. (2008) A female Homo erectus pelvis 36:149–156. Inuits. Am J Phys Anthropol 91:1–20. from Gona, Ethiopia. Science 322:1089–1092.

Weaver PNAS ͉ September 22, 2009 ͉ vol. 106 ͉ no. 38 ͉ 16033 Downloaded by guest on September 26, 2021