sign in sign up
<<
Home , Ardi, Ardipithecus, Chimpanzee, Equatorius, Homo, Oreopithecus

SPECIAL FEATURE: PERSPECTIVE PERSPECTIVE SPECIAL FEATURE: Neither nor , reveals the surprising ancestry of both Tim D. Whitea,1, C. Owen Lovejoyb, Berhane Asfawc, Joshua P. Carlsona, and Gen Suwad,1 aDepartment of Integrative , Human Research Center, University of California, Berkeley, CA 94720; bDepartment of , School of Biomedical Sciences, Kent State University, Kent, OH 44242–0001; cRift Valley Research Service, Addis Ababa, ; and dThe University Museum, The University of Tokyo, Hongo, Bunkyo-ku Tokyo 113-0033, Japan

Edited by Neil H. Shubin, University of Chicago, Chicago, IL, and approved September 10, 2014 (received for review April 25, 2014)

Australopithecus were regularly interpreted during the late 20th century in a framework that used living African , especially , as proxies for the immediate ancestors of the human . Such projection is now largely nullified by the discovery of Ardipithecus. In the context of accumulating evidence from genetics, developmental biology, anatomy, ecology, biogeography, and geology, Ardipithecus alters perspectives on how our earliest ancestors—and our closest living relatives—evolved.

| | hominid | Ethiopia

“...the stock whence two or more have chimpanzees, can serve as adequate repre- (5). Indeed, a widely used textbook still pro- sprung, need in no respect be intermediate sentations of the ancestral past. claims that, “Overall, Au. afarensis seems very between those species.” much like a missing link between the living Background T. H. Huxley, 1860 (1) Africanapesandlaterhomininsinitsdental, ’ Darwin s human evolution scenario attemp- cranial, and skeletal morphology” (6). Charles Darwin famously suggested that ted to explain hominid use, bipedality, Australopithecus can no longer be legiti- was humanity’s most probable birth enlarged , and reduced canine teeth (2). mately viewed as a short-lived transition be- continent, but warned that without fossils, it It easily fit the record of his day, when tween apes and . Rather, it represents was “...useless to speculate on this subject” only a few were known. an adaptive plateau occupied for ∼3Maby (2). Nevertheless, Darwin and his less cau- erectus Austral- was found in the 1890s, and up to four species lineages of small-brained tious contemporaries and intellectual descen- opithecus in the 1920s. Both were rejected as African bipeds. Many assumed that when dants used humans and modern apes to hominids by eminent authorities, but two pre-afarensis fossils were eventually re- triangulate ancestral anatomy and behaviors, grades of human evolution were eventually covered they would increasingly resemble which promulgated the erroneous metaphor Australopithecus recognized. comprised chimpanzees. Today “conventional wisdom” of a hominid “missing link.” Even today, de- different species of small-brained but bi- continues to reflect the deeply held assump- spite thousands of available fossils, this deeply pedal . Homo was its tion that Australopithecus is close to some embedded metaphor reinforces the mis- descendant. imagined chimpanzee-like spe- conceptions that extant apes—particularly In the 1960s molecular data challenged cies. Furthermore, because Australopithecus chimpanzees—can be viewed as “living notions that species lineages of modern apes is often found in open environments, hom- missing links,” or that that modern African and humans could be traced directly to early inid origins are frequently presented as the apes combined can be used to represent the and middle Miocene fossils. The data ulti- tale of a tropical ape forced to adapt to past “as time machines” (3). mately resolved the phylogenetic branching open that expanded via global cli- The notion that modern great apes are order among extant great apes and humans. mate change. The new fossils disrupt such little changed from the last common ances- However, without our current appreciation frameworks. tors we shared with them promoted the of the power of regulatory mechanisms, faith assumption that hominid fossils anatomically in simple DNA similarity also helped reify Conventional Wisdom Challenged intermediate between living apes and our- the notion that chimpanzees were appropriate Ardipithecus is a that ruptures sev- selves would eventually be found. Now, primitive proxies for hominid ancestors (4). eral deeply held perceptions, particularly Australopithecus afarensis however, long sought and recently discovered During the 1970s those visualizing humans as “just a third African fossils provide escape from such discoveries pushed knowledge back to 3.7 persistent but inaccurate projection. These million ago (Ma), but even the iconic paleontological discoveries do not yet include “” differed little from already known Author contributions: T.D.W., C.O.L., B.A., and G.S. designed research; the common ancestor we shared with chim- T.D.W., C.O.L., B.A., J.P.C., and G.S. performed research; T.D.W., C.O.L., South African fossils. The preoccupation B.A., J.P.C., and G.S. analyzed data; and T.D.W., C.O.L., B.A., J.P.C., and panzees (the CLCA). However, they sub- with chimpanzee comparisons led many to G.S. wrote the paper.

stantially reveal the early evolution of the argue that Lucy and her conspecifics walked The authors declare no conflict of interest. “ ” hominid clade (the term hominid denoting like apes, without human-like hip and knee This article is a PNAS Direct Submission. all species on the human side of the human/ extension. Thus, despite a host of unique 1To whom correspondence may be addressed. Email: timwhite@ chimpanzee phylogenetic split). These fossils specializations to committed terrestrial bi- berkeley.edu or [email protected]. “ have begun to rectify the mistaken notion pedality, many declared this species ...very This article contains supporting information online at www.pnas.org/ that contemporary apes, in particular common close to what can be called a ‘missing link’” lookup/suppl/doi:10.1073/pnas.1403659111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1403659111 PNAS | April 21, 2015 | vol. 112 | no. 16 | 4877–4884 Downloaded by guest on September 25, 2021 species of chimpanzee” (7). Broader aspects ramidus skeleton ARA-VP-6/500 (“”) However, the abundant colobine monkeys of Australopithecus paleobiology emerged preserves so many anatomical parts—in such andkudusfoundwithArdipithecus were not gradually during the 20th century. In contrast, clear ecological context—that it transforms adapted to open savannas (as evidenced by the Ardipithecus niche was comprehen- our understanding of early hominid evolution. their microwear, mesowear, isotopes, and sively revealed in ∼250 published pages Ardi preserves crucial elements from a postcranial ecomorphology). By analogous of a single issue of in 2009 (8). single adult female who died 4.4 Ma on a evidence, neither was Ardipithecus,which Perhaps because Ardipithecus was so sud- broad Ethiopian floodplain supporting grassy maintained a -to-forest adaptation denly revealed in so many dimensions of . Her and feet are extraor- well into the Pliocene. Even the earliest context and anatomy—and is so different dinarily preserved. Less well preserved, but Australopithecus species appears to have re- from Australopithecus—aformof“cognitive nevertheless nearly complete elements of her tained elements of woodland adaptation (16). dissonance” settled over some practitioners of teeth, , arms, legs, and provide Accordingly, despite valiant efforts at its re- . The condition’ssymp- further informative anatomy. This fossil and surrection (17), the hypothesis that opening toms range from post hoc cautionary advice associated evidence allow assessment of lo- grasslands led to hominid emergence and (9) to speculation (10) or inexplicable omis- comotion, diet, habitat preference, and even bipedality now stands effectively falsified. sion (11). However, Ardipithecus was one social behavior. of the few surviving lineages of the Miocene The remains of well over 100 additional Locomotion ape adaptive radiation: it preserved funda- individuals from Ardi’s species confirm that Living primates display a wide range of lo- mental arboreal adaptations and it exclu- her critical morphologies are not idiosyn- comotor abilities. Primatologists have for de- “ ” sively shares several independent character cratic but characteristic of the species. Fur- cades attempted to parse these into modes, “ ” complexes with all later hominids. This pri- thermore, these additional individuals reveal such as vertical clinging and leaping, “ ” “ ” mate therefore illuminates human and normal variation for several body parts, such , and . In char- chimpanzee origins in ways that Austral- as the (SI Text, Note 1; Tables S1– acterizations of postural/locomotor adapta- opithecus never could. S4), allowing, for example, assessment of tions of both living and fossil primates, such The Ardipithecus fossils fortuitously ar- . These fossils belong labels are combined with standard postural “ ” rived at an auspicious time in the of to a stratigraphically associated, geologically designations, such as orthograde (upright “ ” biology, just as revelations of developmental contemporaneous, >7,000-specimen assem- trunk posture) and pronograde (largely biology enhanced our understanding of how blage of vertebrate remains identified to horizontal trunk posture). An even more de- morphology evolves. Knock-outs and knock- or species and accompanied by a tailed classification of primate positional and ins, reporter alleles, and other revelations wealth of invertebrate, paleobotanical, and locomotor behaviors partitioned them into about the structural impact of transcription sedimentological data. over 70 categories (18)! The difficulty that such factors, such as homeoboxes and T-boxes pigeonhole terminologies have in encapsulat- (12), have entirely transformed how paleon- A Wooded Habitat ing actual locomotor versatility and diversity tologists can evaluate morphological change. Scenarios about hominids arising in open among living primates is widely recognized. Gould and Lewontin (13) predicted the im- environments go back to Lamarck in African ape locomotion has been intensely pact of such advances even before they ac- 1809 (15). It was widely expected that pre- studied and described. Napier and Napier tually occurred. However, despite ubiquitous Australopithecus hominids would continue to (19) initially assigned chimpanzees (the most citation of their now famous “spandrels” be found associated with open African hab- appropriate body mass analog for Ardipithe- paper (13), adaptationist interpretations of itats. However, the uniquely high-resolution cus) to the category “modified brachiation,” fossil morphology are still the norm in paleo- set of diverse contextual data surrounding the with the trunk usually “held in an upright anthropology. The new understanding of mo- Ar. ramidus remains indicate that Ardipi- position,” and with “true” brachiation used lecular and cellular processes and their roles in thecus preferred wooded habitats that were usually only for short distances. Chimpanzees the transformation of structure (14) promises neither a closed tropical forest nor open are remarkably agile in the trees, even in high to profoundly affect the interpretation of grassland savanna. canopy, despite their large mass. Their sem- hominid fossils. However, as Darwin appreci- The most direct lines of evidence for iorganized “hunts” for monkeys ated, fossils are still the sine qua non in habitat—the “smoking guns” of habitat reveal how remarkably adapted they are to paleobiology. choice—are derived from the Ardipithecus rapid ascent, descent, and navigation across Indeed, uniquely complete fossils such as fossils themselves: (i) craniofacial structure even significant gaps between branches. In Tiktaalik—recovered in accurate chronologi- and masticatory apparatus, (ii) anat- contrast, their is by cal, depositional, ecological, and populational omy and proportions, (iii)macro-andmi- knuckle walking (KW) with occasional bent- contexts—constitute the most effective means croscopic tooth wear, (iv) carbon isotopes in hip-bent-knee (BHBK) bipedality. of illuminating paleobiology. Completeness enamel, and (v) locomotor adaptations. This Every primate species has a diverse posi- and context are critically important. Only primate was adapted to chewing softer, less tional and locomotor repertoire. However, special fossils allow such comprehensive, abrasive foods than any later Australopithecus, the body plan and postcranial anatomy of integrated biological analysis. Historically, and at the same time was much better each largely reflects its dominant and adap- when uniquely complete fossils have been adapted to climbing trees than any other tively crucial locomotor behavior, and con- discovered, their reception has often pre- hominid yet found. strains its range of habitual positional/ cipitated protracted debate. Initial denial is Inferences that early hominids preferred locomotor activities. It is impossible to often followed by “adjustment” in a sort of wooded habitats have been met with skepti- observe such behavioral details for fossil Kuhnian pattern on a smaller scale. Such cism from savannaphils. Soil-based isotopic primates, let alone apply highly parsed cate- progress characterizes the historical sciences data (as well as phytoliths and relatively rare gorizations themselves founded upon the in general, with plate tectonics as a classic fossils of grazing ungulates) do demonstrate more limited array of extant species. example. The partial Pliocene Ardipithecus grass in the species’ regional environment. Nowhere has this problem surfaced more

4878 | www.pnas.org/cgi/doi/10.1073/pnas.1403659111 White et al. Downloaded by guest on September 25, 2021 explicitly than with Ar. ramidus, a species lacked key characters that imbue chimpanzee relative to the upper trunk. This uniquely de- PERSPECTIVE whose postcranium differed dramatically feet with their unique and powerful grasping rived condition places the center of mass over SPECIAL FEATURE: from that of any living primate. Indeed, rel- ability. The modern great ape midfoot is our feet when we stand up, thereby allowing ative to any living or known fossil ape, this foreshortened compared with those of other us to avoid exhaustive BHBK walking. versatile hominid had unique adaptations for primates. The Ardipithecus foot shares short- Although most of her sacrum and lumbar both arboreal and terrestrial locomotion. ened metatarsals with extant African apes, vertebrae were not recovered, parts of both but not the especially shortened distal and lateral halves of Ardi’s pelvis were. The left . Expectations influenced by the phy- mid tarsal bones. As part of this extreme side, although damaged, has several relatively logenetic proximity of humans and chim- shortening of their midfoot, living apes lack complete portions. Importantly, these pre- “ ” panzees led some to posit KW traits in the os peroneum, a sesamoid bone always serve critical landmarks: the auricular surface Australopithecus early (20). Many predicted found in Old World monkeys (as well as (with which the sacrum articulated), the ac- Australopithecus that pre- hands would be living humans). Distinctively, Ardipithecus etabulum (hip joint), and the pubic sym- increasingly chimpanzee-like (21). Living retained this sesamoid and its associated physis (where the two halves of the pelvis African apes have evolved manual adapta- mechanisms. A wider comparison among the meet at its front). Their relative positions in tions for sustained below-branch suspen- soft tissues of humans, chimpanzees, and Ar. ramidus are essentially the same as they sion and regular climbing on varied vertical monkeys reveals that humans also have are in Lucy and other Australopithecus and supports, and for quadrupedal KW. Extant retained many characters lost or modified in Homo;thisreflectsanupperpelvisreor- African ape ray two to five metacarpals chimpanzees as the latter enhanced special- ganized for terrestrial bipedality, characterized are long, especially in chimpanzees. Their izations for pedal grasping (22). by a superoinferiorly (craniocaudally) approx- Ar. ramidus heads bear distinct articular expansion and The foot also shows some imated upper sacrum (sacroiliac articulation) grooving, the latter induced by a lifetime of newly evolved characters that indicate sub- and hip joint, and a broad and more sagittally KW. The primary flexor is reduced stantial adaptation to upright walking. Its facing iliac blade (SI Text, Note 3). and sometimes absent (instead, its tendon great retained its primitive opposability, Even as key upper parts of the Ar. ramidus often joins that to the second ray) (22). The but its lateral rays took on the primary role of pelvis were structurally hominid and clearly central bone (capitate) of the chimpanzee terrestrial propulsion, as shown by a ten- adapted to upright walking, the lower pelvis midcarpal joint demonstrates a smaller range dency for a hypertrophied second ray and an (ischium) that anchors powerful muscles of motion than in Ardipithecus. Chimpanzees Australopithecus-like dorsally developed lat- used in arboreal climbing (adductors and have massive ligaments and structural en- eral metatarsal head and associated mor- hamstrings) was still primitively long, as in hancements of their articular surfaces that phologies. This unique combination indicates pronograde quadrupeds and in climbing firmly and almost rigidly reinforce their palm that Ardipithecus was able to use its lateral apes. This anatomical montage of the pelvis and digits. Ardipithecus, Australopithecus, foot as an effective lever for “toe-off” during uniquely defines Ar. ramidus and its loco- and modern humans lack these. The chim- bipedal progression. motion. Ar. ramidus was less adept in the panzee hand, unlike that of the especially mobile hand and wrist of Ardipithecus,is Lower Back and Pelvis. Equally striking are trees than are living chimpanzees, but was a differences in the lower backs and pelves of more capable climber and clamberer than highly derived. Australopithecus Ardipithecus metacarpals were almost as humans and modern African apes. The latter . Furthermore, its lower short as those of Australopithecus and Homo. have very short lumbar regions. Whereas six back and pelvis bore those fundamental The carpometacarpal joints were primitively to seven are commonly features that provide balance and support less rigid, enabling some palmar flexibility foundinOldWorldmonkeysandknown during bipedal walking. (such as cupping) not seen in chimpanzees. Miocene apes, living apes average only 3.5. Limb Proportions. Thumb musculature appears to have been Moreover, in the modern great ape pelvis, In addition to their spe- cialized hands, feet, and short rigid lower robust, without modification of the tendon’s a strikingly narrowed sacrum combines with attachment to the base of the terminal pollical superiorly extending pelvic “wings” (iliac backs, modern great apes have long arms phalanx, but the pollex itself was shorter and blades) to actually “entrap” and immobilize and short legs that contribute significantly to moregracilethaninhumans.TheArdipi- the lowest lumbar (sometimes two). their remarkable arboreal agility. Human limb thecus This is the case even with inter- and in- proportions are an extreme opposite. The hand is not the anatomical intermedi- ’ ate between humans and modern apes that traspecies variation in degree of superior completeness of the Ardi individual sskele- ’ many had predicted. elongation of the ilium. Lumbar reduction ton (and another individual s associated arm couples with this entrapment “mechanism” bones) allows accurate determination of its Foot. Early anatomists regularly referred to to stiffen the backs in these large-bodied limb proportions. Not only do its ratios differ chimpanzees and as “quadruma- primates, providing injury prevention during from those of both humans and chimpan- nous,” citing their specialized grasping feet. arboreal acrobatics. At the same time, how- zees, they are actually closest to those of most This evolutionary outcome is very different ever, this stiffening makes modern ape trunks known Miocene apes. The latter, with long, fromthehighlyderivedanatomyofthehu- so inflexible that their bipedality requires a flexible vertebral columns and bilaterally nar- foot, which became a shock absorber, BHBK gait. row trunks, have a more primitive locomotor energy store/converter, and locomotor lever. In contrast, humans, earlier Homo,and body plan than does Ardipithecus. However, As with the hand, Ardipithecus foot anatomy Australopithecus retained longer lower backs like Ardi, most of these apes had arms and does not match expectations based on either and evolved uniquely wide sacra. This com- legs of about equal length. This finding fur- chimpanzees or humans. It fits into no living bination allowed all of their lumbar vertebrae ther corroborates that Ar. ramidus was not primate pigeonhole. to participate in a specialized anterior bending dependent on, and perhaps only occasionally Whereas the Ardipithecus great toe was called “lordosis,” in which the lower spine practiced, forelimb suspension as a mode of widely divergent and opposable, its foot actually shifts forward to lie more anteriorly locomotion. Instead Ar. ramidus probably

White et al. PNAS | April 21, 2015 | vol. 112 | no. 16 | 4879 Downloaded by guest on September 25, 2021 preferred slower, more deliberate climbing of pronograde arboreal quadrupeds, but was enamel, enlarged back teeth, reduced canines, and clambering. not as adapted to orthograde vertical climb- and an orthognathous face attached to a short ing or suspension as are extant great apes. cranial base) had evolved from a presumed Arboreal “Multigrady.” We began this Ar. ramidus was neither adapted for quad- primitive condition like that of living chim- section by noting the exceptional difficulty of rupedal running and leaping typical of smaller panzees (4). Surprisingly, the Ardipithecus “classifying” the locomotor and positional pronograde monkeys, nor for the acrobatic skull and dental batteries conform to such behaviors of living primates. Indeed, despite gymnastics of quadrumanous and suspensory- expectations only in limited ways, thereby our best efforts to concisely apply the re- oriented chimpanzees. On the ground, the lack revealing novel evolutionary trajectories for strictive terms widely used in classifying of chimpanzee-like lumbar and pelvic spe- early hominid and modern ape cranio- movement and posture of living primates to cializations allowed Ardipithecus to move dental anatomy. the Ar. ramidus fossils, our efforts have en- with much more effective bipedality than gendered confusion. In hindsight, this is not any living ape. In short, Ar. ramidus com- Dentition and Diets. Ar. ramidus lacks the surprising because the traditional terminol- bined versatile but deliberate climbing specialized dental anatomy reflective of the ogy was never created to accommodate a involving body postures spanning both dietary and feeding specializations of or novel creature such as Ar. ramidus,letalone “pronogrady” and “orthogrady,” with a pre- . Both, but particularly chimpanzees, its more distant ancestors. viously unknown form of bipedality. have broad spatulate associated with Using conventional terminology, we wrote frugivory. Common chimpanzees and bono- that Ar. ramidus “...combined arboreal pal- Terrestrial Bipedality. Despite having re- bos have crowns endowed with pecu- migrade clambering and careful climbing tained hallucial grasping and a primitively liarly thin occlusal enamel in a wide occlusal with a form of terrestrial bipedality...” (8), elongate ischium for climbing, the Ar. ramidus basin suitable for crushing, coupled with and predicted that the CLCA “...[was] upper pelvis was dramatically shortened rel- peripheral crests for shearing. These traits probably a palmigrade quadrupedal arboreal ative to those of all known apes, especially in accord with their dietary preference for ripe climber/clamberer that lacked specializations the distance from its sacral articulation to the fruits and limited folivory. Gorilla molars are for suspension, vertical climbing, or knuckle- hip joint. Along with an inferred long lumbar decidedly higher-cusped and more evenly walking...” (23). region that allowed substantial lordosis, this thin-enameled, resulting in deeper occlusal In the interest of clarity, we will attempt to would have enabled efficient balance control shearing zones reflective of a more fibrous, more plainly spell out the unique aspects of during upright walking with extended hip herbivorous diet. the generalized and highly versatile postural/ and knee, as in Australopithecus and Homo. In these respects, the Ar. ramidus denti- locomotor adaptation of Ar. ramidus.Wedid Differences with these later hominids pri- tion differs from those of modern great apes. not intend to broadly describe Ardipithecus marily relate to the foot, retention of larger It combines somewhat thin molar enamel or the CLCA as quadrupedally -like. posterior thigh muscles, and perhaps sub- (surprisingly thin compared with most Nor did we mean that it displayed an overall tleties of knee and ankle joint kinematics. Australopithecus) with relatively unspecialized locomotor pattern or general postcranial Aspects of currently available ankle mor- incisors and molars. The latter are broadly morphology closely comparable to early Mio- phology hint that the Ar. ramidus knee similar to a range of Miocene apes and to cene apes, such as (6). Ar. ramidus might have been positioned further from Australopithecus, strongly suggesting that go- hadabodyplanthatdifferedfromthesepri- the midline than in later hominids, suggest- rilla and chimpanzee morphologies are each mates, species that were primarily adapted ing slightly greater mediolateral shifts of specialized and independently derived. to pronograde quadrupedality. However, body weight during bipedal progression. Thus, the dietary niches of Ardipithecus Ar. ramidus also exhibited no obvious special- Ar. ramidus also lacked a medial longitudinal and many Miocene apes were almost cer- izations attributable to forelimb dominated pedal arch, necessitating a flat-footed stance tainly broader and less specialized than suspension and vertical climbing, as seen in phase with toe-off from rays two to five. Bi- those of extant African apes. Ar. ramidus the living great apes. Despite retention of pedal progression with extended knee and probably included substantial components of a primitively opposable great toe, its arbo- hip would not have been impeded per se, but opportunistic omnivory, further suggested real locomotion must have been only par- compared with later Australopithecus and by an enamel isotope signature slightly tially African ape-like. Homo, this musculoskeletal architecture different from that of chimpanzees, and Ar. ramidus was fairly large-bodied, so its would have compromised energy dissipa- indistinguishable from those of earliest movements must have been slow relative to tion and loading efficiency in all lower limb Australopithecus. acrobatic chimpanzees, probably usually with joints during walking and running. The Ar. ramidus multiple simultaneous grasps-on-branches implications are that Ar. ramidus was not Cranial Structure. The denti- with both hands and feet. Sometimes ar- as well-adapted to longer, more strenuous tion and craniofacial morphology lack the Ar. ramidus boreally orthograde, was un- bouts of terrestrial bipedality as were known mastication-intensive signals so well estab- Australopithecus doubtedly at other times also effectively Australopithecus species. lished in (e.g., facial size, pronograde, depending upon each chosen ar- zygomatic development, absolute and relative boreal route. Its feeding posture must have Craniodental Evolution postcanine sizes, crown shape, cusp pro- varied extensively. Arboreally, Ar. ramidus is Conventional predictions about stem homi- portions, enamel thickness, wear pattern). best described as a relatively deliberate climber nid craniodental anatomy varied widely be- Combined with the postcranial evidence, this and clamberer, and not as dependent on sus- fore Ardipithecus. Superficially hominid-like aspect strongly suggests a niche considerably pensory-inclined locomotor repertoires as are Miocene ape fossils led many to conclude distinct from that of Australopithecus. living African apes. that Australopithecus evolved from a thick- The Ar. ramidus cranium also differs In summary, Ar. ramidus shared with enameled, robust-jawed Miocene ape (24). substantially from those of the similarly sized African apes and humans a trunk and ap- An alternative idea was that the derived chimpanzee and . By modern-to- pendicular skeleton more evolved than those Australopithecus condition (including thick modern species comparisons, the common

4880 | www.pnas.org/cgi/doi/10.1073/pnas.1403659111 White et al. Downloaded by guest on September 25, 2021 chimpanzee condition is often considered When the iconic Lucy was discovered, it that the small-canined Ardi was male (see PERSPECTIVE less derived than that of the “paedomorphic” was often contrasted with subsequently dis- supplementary materials of ref. 31). SPECIAL FEATURE: bonobo. However, a wider comparison covered, larger conspecifics. Despite greatly The Ardi skeleton, securely established as including Ar. ramidus and gorillas shows expanded sample size, even today the Lucy female, thus also suggests a small degree of that this is true only in a limited sense, and individual, with the most complete post- body size dimorphism in the species, perhaps that both chimpanzees and each cranium of Au. afarensis,isoneofthe similar to that of chimpanzees or humans, as appear to exhibit some uniquely derived smallest members of her species and in many opposed to or gorillas. Appar- features. In particular, the common chim- ways does not represent the modal condition. ently, both male and female Ar. ramidus were ’ panzee s skull has an especially elongate This sampling bias underlies long-standing comparatively larger than many Au. afarensis anterior cranial base (nasopharyngeal re- opposing views. Whereas many observers females, particularly diminutive ones such gion) associated with a more forwardly still argue for a gorilla-like, or even higher level as Lucy. This finding implies that female placed snout and face. This structural pack- of body size dimorphism, such studies are still body size decreased during the emergence of age results in exaggerated prognathism, highly influenced by Lucy, which contributes Australopithecus, perhaps as a part of its a greater gape for its large canines disproportionately to their estimates (29). adaptation to an expanding niche including (25), and a distinctly larger skull relative However, when randomization techniques less wooded habitats. to body size. This anatomical constellation allowing independent and equal contribution is concordant with the species’ inter- and of individuals are used, the results show that Homo sa- intragroup aggressive behavior, especially Social/Reproductive Behavior. Au. afarensis postcranial dimorphism was at piens among males. is the only living primate committed to an intermediate level: larger than in the weakly terrestrial bipedality. We are also the only The anatomically tightly constrained dimorphic chimpanzees and broadly compa- Ar. ramidus living higher primate in which the canine female skull reconstruction rable to modern human levels (∼15% post- plays no social role, and we are the only reveals a cranial capacity as small as that of cranial skeletal dimorphism). These results Austral- primate that engages in prolonged monoga- chimpanzees combined with an stem from a recently doubled available sample opithecus -like short cranial base with derived size now just reaching a statistically appro- mous relationships within the context of a internal flexion. The latter is usually inter- > – larger social group. priate 15 24 individuals (30). Ar. ramidus preted as indicative of bipedality (upright The Ar. ramidus fossils offer an additional, shares the first two of these trunk and neck posture) and less often linked new, and unexpected perspective. Ardi’slimb characters with humans, which may elucidate with altered structure. However, in bones are quite large, with measures ex- the third. All of the earliest known hominids Ardipithecus features commonly related to posture, such as ceedingthoseofLucybyasmuchas30%. ( , ,and ) the position of the , the Furthermore, the largest of eight known had apparently already abandoned the difference from a gracile-faced ape such as “ ” Ar. ramidus humeri is only slightly larger primitive C/P3 honing complex in which a the bonobo is slight, casting significant doubt than Ardi’s estimated humeral size, and is triangular, projecting upper canine is contin- on that functional explanation. Rather, some comparable to those of the larger, pre- uously sharpened by occlusion against the subtle changes between internal (neural) and sumably male specimens of Au. afarensis. anterior lower , especially in males. external (cervico-pharyngeal) developmental Importantly, Ardi’s completeness allows a This phenomenon is ubiquitous among living parameters may have altered cranial base rarely possible craniodental sex assessment proportions and flexion as a simple non- and fossil apes and monkeys. However, even of her postcranium independent of infer- Ar. ramidus adaptive corollary of other craniofacial struc- the male canines of are feminized: ences based on body size. tural changes. Compared with Ar. ramidus they are short and morphologically blunt, The relevant craniodental indicators of and Sahelanthropus,earlyAustralopithecus with tips wearing down to the level of the species had, on average, up to perhaps 25% dimorphism in Ardi are thus her very small surrounding teeth. larger cranial capacity. Perhaps this was linked canines and gracile facial skeleton (especially It has been argued recently that because thin supraorbital torus), suggesting a female of “trade-offs” between canine projection with altered behavior, sociality, and cognition Ar. ramidus assignment. A male might have (and resulting gape) and masticatory muscle accompanying niche expansion (26) and Sahelanthropus transition to more fully committed terres- looked more like the cranium function,malecaninesizereductionmight triality (e.g., ref. 27). that shows basicranial and facial structural have been associated with chewing function patterns similar to Ardi, but is also endowed in early hominids (25). This argument, ap- with a thick browridge, taller face, and de- Sexual Dimorphism. One of the strongest plied to emerging hominids, is dubious veloped nuchal crest. Nevertheless, we ini- and most persistent misconceptions about early because Ar. ramidus shows dramatic male tially considered the possibility that Ardi hominid paleobiology concerns the almost canine height reduction but no obvious signs Australopithecus might be a small-canined, gracile-headed universal acceptance that of masticatory enhancement. It is therefore (and in particular Au. afarensis) exhibits male. We therefore paid particular attention far more likely that reduction of male canine large, gorilla-like body size sexual dimorphism to addressing the sex of Ardi on probabilistic size and height, especially of the upper ca- (in which linear skeletal dimensions differ by grounds. By modeling canine dimorphism ∼30% between sexes, and males have ∼2× across the entire Aramis sample (repre- nine, signals a fundamental change in social female body mass). To some observers, sented by up to 21 individuals), we found behavior. Moreover, bipedality and male this has implied strong levels of male–male Ar. ramidus canine size dimorphism to be canine feminization appear to have been competition. However, because exaggerated only slightly greater (average canine dimen- evolutionarily coupled. Just as its hands, body size dimorphism is usually coupled with sions ∼12%largerinmalesthaninfemales) feet, and pelvis indicate a unique locomotor strong canine dimorphism among higher than the modern human condition. This small behavior and ecological role for Ar. ram- primates (but not in Australopithecus), this degree of dimorphism is nevertheless, in turn, idus, so too do its face and teeth suggest that has created a conundrum (28). sufficient to effectively preclude the possibility stem hominid social behavior was novel.

White et al. PNAS | April 21, 2015 | vol. 112 | no. 16 | 4881 Downloaded by guest on September 25, 2021 Placing Ardipithecus in Life’s Tree Ar. ramidus obviously postdates the CLCA. It does not exhibit the derived features charac- teristic of either chimpanzees or gorillas. In contrast, Ar. ramidus shares four ma- jor, newly evolved, morphogenetically in- dependent character complexes with later Australopithecus (as judged by comparison with fossil and modern outgroup apes): (i) a nonhoning C/P3 complex and feminized male canine (31), (ii) a short, broad cranial base (32), (iii) a broadened, shortened and “twisted” ilium (SI Text, Note 3), and (iv) tarsal and metatarsal/phalangeal special- izations related to upright walking. No other fossil or modern ape shares such evolu- tionary derivations. It is therefore likely that Ardipithecus lies within the hominid clade (Fig. 1 and Fig. S1). At 4.4 Ma, it is obviously not an ancestral chimpanzee. Nor is it likely to be a relict ape that bears no special re- lationship to hominids (9, 10). Continuity with Australopithecus How is Ardipithecus related to later hominids? Fig. 1. Evolution in different directions. Pan troglodytes (left boxes); Ar. ramidus (center boxes); H. sapiens (right The temporally nearest Australopithecus is the boxes). Micro-CT renders on blue backgrounds. Ar. ramidus nests in the hominid clade based on uniquely derived Au. anamensis character complexes shared exclusively with Australopithecus and Homo. In the skull (A), chimpanzees combine little-known taxon . Recent a primitively long posterior, and derived elongate anterior basicranium; hominids share derived shortened bases. In the data from carbon isotopes of enamel (16), dentition (B), hominids lost the primitive functional honing complex of fossil and modern great apes based on associated flora and fauna (33), and cranio- a projecting male canine (male Ar. ramidus dentition, ARA-VP-1/300, shown; there is no honing). In limb proportions dental anatomy suggest that this species was (C), Ar. ramidus is primitive relative to the derived elongated arm of the chimpanzee and the lengthened leg of the ’ related to Ardipithecus either as a close col- human. Radius and tibia outlines adjusted to mean lengths; ARA-VP-6/500 tibia s missing distal end is conservatively restored per SI Text, Note 2. In the pelvis (D), hominids share broader, lower iliac blades (yellow arrows; white brackets lateral relative, or in an ancestral-descendant show the superior and inferior extent of the sacrum’s articular surface; short blue line on fossil indicates superior relationship. The two taxa are superimposed margin of hip joint). Note the low position of the sacral joint in hominids. The chimpanzee hand (E) bears elongate in our study area’s single stratigraphic se- metacarpals and stiffening for climbing and suspension. Hominids primitively retain short metacarpals, whereas quence, but an inadequate fossil record there humans have shortened phalanges. The Ar. ramidus foot (F) shows a primitive midfoot that had not evolved into the shortened flexible structure of living apes; neither had it evolved into a modern human-like foot that functions both as and elsewhere makes it premature to choose a stiffened lever and a compliant shock absorber. among three different phylogenetic hypothe- ses (8), although the Burtele foot (34) suggests that some Ardipithecus species may have Ar. ramidus—Au. anamensis—Au. afarensis, Ape and Human Evolution: Deciphering persisted past 3.5 Ma. and beyond. This is additional evidence the Hominoid Bauplan Despite dramatic morphological distinctions that Australopithecus is the known taxon All higher vertebrates have individual overall between Ar. ramidus and Australopithecus phylogenetically most closely related to body plans (or bauplans forhighertaxa)that that signal niche differentiation, Ar. ramidus Ar. ramidus. Although the cranium is repre- constitute a sort of fundamental structural shares many finer details of structure and sented by many fewer specimens than the “blueprint.” Evolutionary reorganization of a Australopithecus morphology with early . dentition, it is nevertheless similarly revealing. bauplan almost always involves deeply struc- This aspect is particularly true of the more As with the dentition, the cranial base shows tured, concurrent changes in many characters. bauplan abundant dental remains representing that Ardipithecus and Australopithecus share Albeit at a smaller scale, the meta- many individuals of both taxa. These remains so much structural detail that the likelihood of phor helps to conceptualize evolution in the Ardipithecus allow assessments of variation, as well as parallel acquisition is de facto negated (32, in great ape and human clade. The the discernment of temporally sequenced fossils reveal an unexpected primate with a contrast with ref. 9). morphoclines between Ardipithecus and unique bauplan. Such morphological sequences show that Australopithecus, and indeed among chro- Ardipithecus shares some key evolutionary distinct adaptive complexes, such as those nospecies of the latter. novelties with all living apes and hominids exemplified by Ar. ramidus and Austral- For example, it is not just the itemized opithecus (fossil and modern). Prominent among these character state of the canine, as “small” (or , probably arose through a complex is the anterior migration (“invagination”)of “apically wearing”) that exclusively aligns the sequence of evolutionary modifications. Some the deeply into the thoracic Ar. ramidus anatomical components (particularly in the C/P3 complex with later homi- space. This Miocene acquisition probably nids. Rather, the entirety of this morpho- dentition) probably emerged gradually in occurred independently multiple times (35, logical complex, comprising a multitude of terms of geological time, whereas other mor- 36), transforming the primitive circular cross- largely (or partially) independent features, pho-functional complexes such as those of the section of the original ape trunk (today only show obvious phenetic continuity and foot and pelvis may have changed more retained in monkeys) into the flattened, gradations along the series: Ar. kadabba— abruptly under stronger selective pressures. elliptical one in our ancestors and ape

4882 | www.pnas.org/cgi/doi/10.1073/pnas.1403659111 White et al. Downloaded by guest on September 25, 2021 relatives. This shift positioned the shoulder For example, although tailless, hand that characterize today’s African (or PERSPECTIVE joint more posteriorly and laterally, allowing and Proconsul were primitive, exhibiting qua- Asian) apes. Indeed, whereas we hypothesize SPECIAL FEATURE: greater ability to circumduct the arm during drupedal pronograde bauplans similar to those that the ancestor (CLCA) was not a quadru- arboreal climbing and clambering. Other of some Old and New World monkeys. pedal primitive ape sensu any so-far known forelimb structures associated with weight combined a largely Proconsul- Miocene genus, neither was it as specialized as bearing during pronograde postures became like body plan with special climbing adapta- today’s highly derived African relicts of the modified. tions including proportionally large forelimbs Miocene ape radiation. This transformation was interpreted in the (35). is reported to exhibit Although obviously not the common an- 19th and 20th centuries to be the result of changes in its wrist similar to those we share cestor of humans and chimpanzees, Ardipi- natural selection favoring below-branch sus- with living great apes, but had only a partially thecus nevertheless provides strong evidence pension and brachiation, or varied forms of invaginated vertebral column and short palms with which to infer that the CLCA was climbing with the trunk in vertical or near- and dorsiflexing fingers (36). a generalized African ape. It was probably a vertical postures (i.e., the suspensory ortho- exhibits suspensory features in its forelimb relatively large-bodied primate with a unique, grade bauplan). According to this perception, includinglong(butalsodorsiflexing)fingers. previously unknown bauplan combining in- the spinal and thoracic reorganizations (see Another genus, ,wasahighly vaginated vertebrae within a flattened thorax, earlier) were adaptations for orthograde body specialized folivore exhibiting extreme adap- enhanced shoulder mobility, elbow extension posture and locomotion, as were enhanced tations to suspension, with forelimbs of close to that in living apes, and ulnar with- forelimb length and shoulder mobility for uniquely exaggerated length among known drawal at the wrist. However, it otherwise vertical climbing and suspension. However, Mioceneapes.Itsuniqueanatomyissope- lacked subsequently elongated forelimbs/fore- Ardipithecus, lacking forelimb elongation culiar that it is now universally viewed as arms/hands and other specializations related and other key suspension-linked features, a specialized, extinct insular form. Oddly, to habitual suspension and enhanced vertical shows that vertebral column “invag- however, over the decades (38) and con- climbing. The CLCA foot maintained many ” tinuing today (39), real and imagined ele- primitive structures related to plantar- ination was not adaptively linked with Oreopithecus either orthograde climbing or suspension ments of anatomy have been flexive propulsion, but probably had also per se. Rather, the derived thoraco-vertebral mistaken as conferring hominid status and evolved plantigrade features, such as heel- used as “cautionary tales” about the ability of modifications are best considered a conse- strike. However, it still lacked the peculiar parallelism to confound phylogenetic signal. quence of an adaptive response for en- substrate-conforming, hand-like grasping For example, it was recently erroneously hanced forelimb circumduction. This foot of living African apes. As one of the claimed (9) that Oreopithecus exhibits in- ’ Ardipithecus bauplan shift must have occurred as a part CLCA s descendants, was an dependently evolved specializations (“homo- of increasingly versatile climbing and clam- early biped that lacked the derivations that plasies”)withArdipithecus. Fig. S1 illustrates chimpanzees would evolve, yet had not yet bering adaptations, but with much less em- the flaws of such claims. acquired specializations for a greater com- phasis on suspension and vertical climbing Extensive parallelism is well-documented mitment to terrestrial bipedality (such as than previously thought. among higher primates. As the architects of hallucial adduction and shortened ischia). The deeply embedded perception of the “ ” bauplan the Modern Synthesis and their intellectual traditional suspensory orthograde forbears recognized long before the genetic Living Apes: The Pitfalls of Projection has led many to project the anatomical and developmental principles known today, The last half-century of field and laboratory packages exemplified by living apes onto identifying -based shared derived research on our African ape relatives has hypothesized common ancestors (4, 11). characters is a challenge in paleobiology. revealed a great deal about their biology. Ardipithecus However, lacks the special- However, with multiple, independent Darwin was clear on the pitfalls of projecting izations of living African (and Asian) apes, character complexes exclusively linking living primates into the past as proxy such as the long forelimbs and short, stiff Ar. ramidus to later hominids, cautionary ancestors, but without a fossil record these back. Its feet and hands do not indicate re- tales about parallel evolution based on cautions proved difficult for many to heed. liance on suspension or extant ape-like ver- misinterpretations of exemplars such as As Moore (40) has noted, using chimpanzee tical climbing. A recent limited study on the Oreopithecus seem diversionary rather than anatomy, physiology, behavior, and ecology ∼ proximal of the 6 Ma hominid constructive. to model early hominids is useful for care- Orrorin and a range of Miocene apes also fully delimited modeling exercises, but simply concluded that the living Asian and African The Last Common Ancestor. Despite a projecting living forms into the past as apes are probably independently derived clear phylogenetic nesting of hominids proxiesdoesnotqualifyassuch:“...their from a more generalized ancestor (37). We with African apes, and despite a plethora greatest danger is that models are easily predict that similar corroborative conclusions of middle and forms now mistaken for that which they model; critically, will follow detailed analyses of other ana- known, the CLCA remains paleontologically by creating a mental prototype they can mask tomical regions. elusive. Nevertheless, Ar. ramidus predicts variation and so block use of the comparative much about its structure, and illuminates method for testing hypotheses...This danger Miocene Apes. The markedly diverse fossil the emergence of evolutionary novelty in is especially serious in hominid modeling.” apes of the Miocene (currently ∼20 to >25 the human, chimpanzee, and gorilla . Now that there is an emergent fossil record known genera) play no small role in un- As inferred from Ardipithecus and other of hominid origins, to what extent do the derstandingtheevolutionarysignificanceof hominid fossils, the ancestral bauplan of living apes, particularly chimpanzees, help Ar. ramidus.Someofthese∼23–7Maforms the CLCA seems not to have involved us understand our evolution ...and theirs? are known to have unique bauplans that vary a short and stiff lower spine, especially Not surprisingly, our closest living rela- from primitive to partially derived in terms of elongate forelimbs, or the many KW/ tives turn out to be important not just for features discussed above. suspensory adaptations of the wrist and their value in biodiversity, behavioral, and

White et al. PNAS | April 21, 2015 | vol. 112 | no. 16 | 4883 Downloaded by guest on September 25, 2021 physiological studies, but also as models a similarly generalized African ape ancestor ditionally, Ar. ramidus provides only limited and case examples regarding specifics of was the gorilla. In contrast to chimpanzees, evidence about the and timing of evolutionary ecology and patterning. Chim- gorillas characteristically occupy a larger-bod- crucial early events in hominid evolution. panzees are also highly informative hominid ied herbivory-frugivory niche, with smaller However, even this evidence is important in comparators precisely because they turn out home ranges that enable pronounced sexual removing the confinements of the missing- to be so different from modern people and dimorphism and one-male, multiple-female link mentality that have distorted inter- fossil hominids. social groups. pretations of human evolution for more Ar. ramidus was a primate with no close Hominids appear to have emerged by de- than a century. In this manner, a distant analog among living monkeys or apes. Before veloping a search-intensive terrestrial feed- Pliocene primate has again demonstrated its discovery, conventional wisdom held that ing niche, accompanied perhaps by food ’s unique contribution to un- chimpanzees were largely primitive, and that transport and sharing in less densely for- derstanding the on earth. humans were derived from forms much like ested but still wooded areas. In this man- them. That meant that the degree of “prim- ner, Ardipithecus represents an evolutionary, ACKNOWLEDGMENTS. We thank the Ministry of Cul- ” ture and Tourism, the Authority for Research and Con- itiveness of characteristics (including be- adaptive, ecological, and anatomical bridge servation of Cultural Heritage, the National Museum of havior) could be determined through simple between Australopithecus and the yet to Ethiopia, and the Afar Regional Government, Ethiopia for permissions and facilitations; the Afar people of the Mid- comparison with the chimpanzee homolog. be found late Miocene African ape that was dle Awash and many other field and laboratory workers The fallacy of this logic is amply illustrated by commonly ancestral to both chimpanzees who contributed directly to the research efforts and Ardipithecus. For example, parsimony-based results since 1981; and William Kimbel and three anon- and ourselves. ymous reviewers for helpful comments on the manu- appeals to KW in contemporary African apes Unfortunately, the still sparse fossil re- script. This work was supported in part by the Institute have been used to argue that this locomotor cord only exposes part of the locomotor of Geophysics and Planetary Physics of the University of mode must have been the primitive condi- bauplan California at Los Alamos National Laboratory, the Japan diversity and transitions during Society for the Promotion of Science, and the National tion for our last common ancestors with the Miocene ape adaptive radiation. Ad- Science Foundation. them (3, 4). However, despite intensive searching of African, European, and Asian deposits, no compelling Miocene evidence 1 Huxley TH (1860) The Origin of Species. Collected Essays Vol. 2 23 Lovejoy CO, Suwa G, Simpson SW, Matternes JH, White TD (2009) – of KW has so far been found; Ar. ramidus Darwiniana (Macmillan, London), pp 71 79. The great divides: Ardipithecus ramidus reveals the postcrania of our 2 Darwin C (1871) The Descent of Man (John Murray, London). last common ancestors with African apes. Science 326(5949):100–106. strongly suggests that none will be, at least in 3 Wrangham R, Pilbeam D (2001) African apes as time 24 Andrews P, Martin L (1991) Hominoid dietary evolution. Philos any candidates for the last common ancestor. machines. All Apes Great and Small Volume I: African Apes,eds Trans R Soc Lond B Biol Sci 334(1270):199–209, discussion 209. Galdikas B, Briggs E, Shapiro S, Goodall J (Kluwer Academic/ 25 Hylander WL (2013) Functional links between canine height and – Plenum, New York), pp 5 17. jaw gape in catarrhines with special reference to early hominins. Am J Fossils: Evolutionary Pathways Uniquely 4 Pilbeam D (1996) Genetic and morphological records of the Phys Anthropol 150(2):247–259. Revealed Hominoidea and hominid origins: A synthesis. Mol Phylogenet Evol 26 White TD, et al. (2006) Asa Issie, Aramis and the origin of 5(1):155–168. Australopithecus. Nature 440(7086):883–889. In focusing on human evolution, it is rarely 5 Stern JT, Jr, Susman RL (1983) The locomotor anatomy of 27 Ward CV, Kimbel WH, Johanson DC (2011) Complete fourth Australopithecus afarensis. Am J Phys Anthropol 60(3):279–317. mentioned that hominid fossils serve as in- metatarsal and arches in the foot of Australopithecus afarensis. 6 Fleagle JG (2013) Primate Evolution and Adaptation (3 Ed) Science 331(6018):750–753. formative outgroups for evaluating African (Academic, San Diego). 28 Plavcan JM, van Schaik CP (1997) Interpreting hominid behavior apes. This is because Australopithecus (and 7 Diamond J (1991) The Rise and Fall of on the basis of sexual dimorphism. J Hum Evol 32(4):345–374. Homo (Radius, London). even more so ) is derived in so many 29 Gordon AD, Green DJ, Richmond BG (2008) Strong postcranial 8 White TD, et al. (2009) Ardipithecus ramidus and the paleobiology size dimorphism in Australopithecus afarensis: Results from two new ways. However, with the chronological and of early hominids. Science 326(5949):75–86. evolutionary depth that Ar. ramidus now 9 Wood B, Harrison T (2011) The evolutionary context of the first resampling methods for multivariate data sets with missing data. – hominins. Nature 470(7334):347–352. Am J Phys Anthropol 135(3):311 328. contributes, contrasting it with extant African 30 10 Harrison T (2010) Anthropology. Apes among the tangled Reno PL, McCollum MA, Meindl RS, Lovejoy CO (2010) An apes reveals that major aspects of living branches of human origins. Science 327(5965):532–534. enlarged postcranial sample confirms Australopithecus afarensis African ape dietary, locomotor, and socio- 11 Thorpe SK, McClymont JM, Crompton RH (2014) The dimorphism was similar to modern humans. Philos Trans R Soc Lond – behavioral adaptations must have evolved arboreal origins of human . Antiquity 88(341): B Biol Sci 365(1556):3355 3363. 906–914. 31 Suwa G, et al. (2009) Paleobiological implications of the after their splits with the hominid clade. With 12 Carroll SB, Grenier JD, Weatherbee SD (2005) From DNA to Ardipithecus ramidus dentition. Science 326(5949):94–99. the new hominid fossils, it is now possible to Diversity: Molecular Genetics and the Evolution of Design 32 Kimbel WH, Suwa G, Asfaw B, Rak Y, White TD (2014) better appreciate chimpanzee specializations (Blackwell, London), 2nd Ed. Ardipithecus ramidus and the evolution of the human cranial base. 13 Gould SJ, Lewontin RC (1979) The spandrels of San Marco and Proc Natl Acad Sci USA 111(3):948–953. in anatomy, locomotion, diet, and aspects of the Panglossian paradigm: A critique of the adaptationist 33 Geraads B, Bobe R, Manthi FK (2013) New ruminants (Mammalia) programme. Proc R Soc Lond B Biol Sci 205(1161):581–598. from the Pliocene of Kanapoi, , and a revision of previous social behavior. The latter include species 14 Shubin N, Tabin C, Carroll S (2009) Deep homology and the collections, with a note on the Suidae. J Afr Earth Sci 85:53–61. characteristics of chimpanzees, such as terri- – origins of evolutionary novelty. Nature 457(7231):818 823. 34 Haile-Selassie Y, et al. (2012) A new hominin foot from Ethiopia 15 toriality and intergroup , complex Lamarck JB (1809) Philosophie Zoologique (Dentu, Paris). shows multiple Pliocene bipedal adaptations. Nature 483(7391):565–569. 16 Cerling TE, et al. (2013) Stable isotope-based diet reconstructions 35 Nakatsukasa M, Kunimatsu Y (2009) Nacholapithecus and its male alliances, strong intragroup competition of Turkana Basin hominins. Proc Natl Acad Sci USA 110(26): importance for understanding hominoid evolution. Evol Anthropol and aggression linked to “advertised” female 10501–10506. 18(3):103–119. 17 Domínguez-Rodrigo M (2014) Is the “savanna hypothesis” estrus, and unusually robust sperm compe- 36 Moyà-Solà S, Köhler M, Alba DM, Casanovas-Vilar I, Galindo J a dead concept for explaining the emergence of the earliest (2004) Pierolapithecus catalaunicus, a new Middle Miocene great tition resulting from extreme female pro- hominins? Curr Anthropol 55(1):59–81. ape from . Science 306(5700):1339–1344. miscuity. Accompanying physiologic and 18 Hunt K, et al. (1996) Standardized descriptions of primate 37 Almécija S, et al. (2013) The femur of Orrorin tugenensis exhibits locomotor and postural modes. Primates 37(4):363–387. genetic novelties support the idea that the morphometric affinities with both Miocene apes and later hominins. 19 Napier JR, Napier PH (1976) A Handbook of Living Primates chimpanzee adaptive complex is derived. (2 Ed) (Academic, San Diego). Nat Commun 4:2888. 38 “ ” 20 Richmond BG, Strait DS (2000) Evidence that humans Hürzeler J (1958) Oreopithecus bambolii Gervais: A preliminary Rather than primitive living missing links, – evolved from a knuckle-walking ancestor. Nature 404(6776): report. Verh Naturf Ges. Basel 69(1):1 48. chimpanzees are thus disclosed as specialists 39 382–385. Köhler M, Moyà-Solà S (1997) Ape-like or hominid-like? The that probably initially evolved in tropical 21 Tocheri MW, Orr CM, Jacofsky MC, Marzke MW (2008) The positional behavior of Oreopithecus bambolii reconsidered. Proc Natl , most likely when such habitats went evolutionary history of the hominin hand since the last common Acad Sci USA 94(21):11747–11750. ancestor of Pan and Homo. J Anat 212(4):544–562. 40 Moore J (1996) Savanna chimpanzees, referential models and the through episodes of recurrent fragmentation 22 Straus WL, Jr (1949) The riddle of man’s ancestry. Q Rev Biol last common ancestor. Great Ape Societies,edsMcGrewWC, and expansion. An earlier derivation from 24(3):200–223. Marchant LF, Nishida T (Cambridge Univ Press, Cambridge), pp 275–292.

4884 | www.pnas.org/cgi/doi/10.1073/pnas.1403659111 White et al. Downloaded by guest on September 25, 2021