HAL Id: halshs-00428700https://halshs.archives-ouvertes.fr/halshs-00428700

Submitted on 29 Oct 2009

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Were climatic changes a driving force in hominidevolution?

Jean Chaline, Alain Durand, Anne Dambricourt Malassé, Bruno David,Françoise Magniez Jannin, Didier Marchand

To cite this version:Jean Chaline, Alain Durand, Anne Dambricourt Malassé, Bruno David, Françoise Magniez Jannin,et al.. Were climatic changes a driving force in hominid evolution?. Climates : Past and Present,Geological Society, pp.185-198, 2000, Geological society. <halshs-00428700>

J. CHALINE1, A. DURAND1, A. DAMBRICOURT MALASSÉ2, B. DAVID 1, F.MAGNIEZ-JANNIN1 & D. MARCHAND1

1 UMR CNRS 5561 et Laboratoire de Préhistoire et Paléoécologie du Quaternaire del'EPHE, Université de Bourgogne, Centre des Sciences de la Terre, 6 bd. Gabriel, 21000

Dijon, France2 Institut de Paléontologie humaine du Muséum National d'Histoire Naturelle de Paris

(UMR CNRS 9948), 1 rue R. Panhard, 75013 Paris, France(e-mail: jchaline@ u-bourgogne.fr)

Abstract: A comparison of externalist and internalist approaches in hominid évolutionshows that thé externalist approach, with ils claim that climate was responsible for théappearance of bipedalism and hominization, now seems to be ruled out by thé biological,palaeogeographical, palaeontological and palaeoclimatic data on which it was based.Biological data support thé embryonic origin of cranio-facial contraction, which determinedthé increase in cranial capacity and thé shift in thé position of Iheforamen magnum implyingbipedalism. In thé internalist approach, developmental biology appears as thé driving forceof hominid évolution, although climate exerts a significant influence and was involved in théfollowing ways: (1) in thé prior establishment of ecological niches that allowed thé commonancestor to become differentiated into three subspecies; (2) by dividing up thé area ofdistribution of species, resulting in thé present-day subspecies of gorillas and chimpanzees;(3) by facilitating relative fluctuations of thé geographical areas of distribution of thévarious species, particularly thé spread of australopithecines across thé African savannafrom north (Chad, Ethiopia) to south (South Africa); (4) by determinîng adaptivegeographical difîerentiations among Homo erectus and Homo sapiens (pigmentation,haemoglobin, etc.).

Climate, as a factor of thé environment, exerts amajor influence over thé geographical distribu-tion of species and subspecies, and over theiradaptation to their environment, although thereare wide variations in thé way it affects différentgroups. The living world is organized on ahierarchical basis, and thé gearing ratios be-tween thé différent levels may vary substantially.This is thé case with hominids, where changesare small at thé molecular level (King & Wilson1975; Hixson & Brown 1986; Miyamoto et al.1987, 1988; Hayasaka et al. 1988; Ueda et al.1989; Gonzales et al. 1990; Sibley et al. 1990;Bailey et al. 1991, 1992; Horai et al. 1992;Perrin-Pecontal et al. 1992; Goodman et al.1994), larger at thé chromosomal level (Chiarelli1962; De Grouchy et al. 1972; Stanyon &Chiarelli 1981, 1982, 1983; Yunish & Prakash1982; Marks 1985, 1993; Dutrillaux & Couturier1986; Dutrillaux et al. 1986; Godfrey & Marks

1991; Matera & Marks 1993) and greatlyamplified at thé morphological level (Dambri-court Malassé 1987, 1988, 1993, 1996; Chaline1994, 1998; Deshayes, 1997). Thèse différencesconstitute thé human paradox.

The questions raised are about thé précise rôleof climate in distinct aspects of hominid évolu-tion and thé way climate affects thé différentlevels of organization of thé living world. Thereare currently two contrasting views of thé originand évolution of thé higher apes and ofhumankind as related to thé rôle of climate.One is an externalist approach (Coppens 1986,1994), whereas thé other is an internalistapproach (Chaline 1994, 1998; Chaline et al.1998).

The aim of this paper is: (1) to présent théexternalist and internalist approaches; (2) to putforward explanatory models of évolution withina paiaeoclimatic and palaeoecological frame-

From: HART, M. B. (éd.) Climates: Past and Présent. Geological Society, London, SpécialPublications, 181, 185-198, 1-86239-075-4/S15.00

186 J. CHALINE

work; (3) to test thé exact rôle of climate as adriving force behind human évolution (when,where and how?), and thus to test thé twoopposing hypothèses, thé externalist v. théinternalist approach.

The externalist approach: climate as thémajor drivin g force behind human évolution

The externalist approach (Coppens 1986, 1994)is a theory in which climate is seen as théprincipal driving force behind human évolution,alongside tectonics. The theory is based on twoobservations: (1) thé present-day distribution ofchimpanzees and gorillas covers ail of thétropical forest régions of Africa, but terminâteswith virtually no overlap at thé gréât tectonictrough of thé East African Rift Valley; (2) thésites of hominid fossils that are more than 3 Maold lie to thé east of thé Rift Valley.

Coppens deduced from this that panids andhominids were segregated geographically by thétectonic crisis some 8 Ma ago that caused théRift Valley floor to collapse and raised théwestern lip of thé Western Rift, creating an8000m barrier between thé summits and thébottom of Lake Tanganyika. 'The gréât initialprovince was split in two. The west remainedhumid and thé forests and woodlands werepreserved, while thé east became increasinglyarid with ever more sparse savanna. Thecommon ancestor population of Panids andHominids was therefore divided into a largerwestern and a smaller eastern population. It istempting to imagine that thé ségrégation was thécause of thé divergence between thé two groups:thé western descendants continued to adapt tolif e in an arboreal environment and formed théPanids while thé eastern descendants of thé sameancestors "invented" new mechanisms suitablefor lif e in an open environment and gave rise tothé Hominids' (Coppens 1994, p. 67). Coppensconsidered that thé earliest forms of australo-pithecines were more confined to arborealenvironments than thé more récent species(Coppens 1986, p. 232). Man, for his part, iswithout doubt a pure product of aridity, whichCoppens termed thé 'Omo event' (Coppens1994, p. 71).

This theory asserts that thé diversity currentlyobserved among hominids results from tectonicand subsequently climatic events. Climate is seentherefore as a major force in human évolution.The gréât weakness of thé externalist approachis that it provides no biological mechanism toaccount for thé substantial changes in bodyshape.

The internalist approach: thé ontogeneticmechanism

This is a radically différent theory based uponbiological data, especially upon thé ontogeneticmechanism. This theory was later explored byDelattre & Fenart (1954, 1956, 1960), whoanalysed thé ontogenesis of higher ape andhuman skulls and showed that they underwentoccipital and facial contraction. Thèse majorstages of cranio-facial contraction are a well-known phenomenon in primates, alternativelytermed 'flexure of thé skull base' or 'occipitalshift' (Deniker 1885; Anthony 1952).

The internalist approach is supported by thédata of ontogenetic development, which wascompletely overlooked in thé so-called SyntheticTheory of Evolution during thé 1940s (Devillers& Chaline 1993). Such an approach linksgenetics to morphogenesis through developmen-tal biology. It has given rise to stimulatingresearch (Gould 1977; Raff & KaufTman 1983;Devillers & Chaline 1993; Raff 1996 (forsynthesis); McGinnis & Kuziora 1997; Pennisi& Roush 1997).

This approach has been taken up morerecently by Deshayes (1986, 1991), Dambri-court Malassé (1987, 1988, 1996) and Deshayes& Dambricourt Malassé (1990). DambricourtMalassé (1987) discovered and proved thatcranio-facial contraction is a embryonic phe-nomenon that is clearly visible in thé mandible,and that thé living and fossil lower jaws retainthé range of contraction. She also demonstratedthat primate évolution can be summarized as six'fundamental ontogenies', because of their em-bryological origin, corresponding to successivebody plans or bauplans, resulting from six majorphases of cranio-facial contraction: (1) prosi-mians; (2) monkeys apes; (3) gréât apes; (4)australopithecines; (5) Homo', (6) Sapiens. Ré-cent morphometric studies on thé skull (Chalineet al. 1998) hâve shown that thé modem humanform is not significantly différent from thé earlierforms of thé genus Homo. Thus we do not retainhère thé 'Sapiens' bauplan (Fig, 1).

The morphological areas defined by thèsebauplans are exploited by more-or-less numer-ous speciation events (Fig. 2). Obviously,comparisons between thé gréât ape skull plan,which is contemporary with Homo sapiens, andthé two fossil hominids skull plans (Australo-pithecus and Homo), which extend over time, area working hypothesis and not an evolutionarymodel in thé strict sensé, insofar as we assimilatethé living gréât apes to their Tertiary ancestors.

This internalist approach views development

CLIMATI C CHANGES AND HOMINID EVOLUTION 187

PROSIMIANS MONKE Y GREATAPES AUSTRALO-

Fig. 1. Phylogeny of embryonic basicranial-facial contraction in primates (after Dambricourt Malassé 1996). X,range of contraction; Y, geological time. 1, Adapiform; 1', Lemuriform; 2, Mesopithecus; 2', Cercopithecus; 3,Proconsul africanus; 3', Gorilla; 4, Australopilhecus africanus; 5, Homo erectus; 6, Homo sapiens. T, extinction.Explanation: thé range of embryonic contraction is unchanged from Adapiforms (1) to living lemurs (!' includinglorises and tarsiers). However, it increases between Adapiforms and Homo. Each column represents onefundamental ontogeny. Diversity occurs in each: in terms of range Tarsius fits in with thé prosimians and not thésimians, and similarly Hylobates fits in with thé simians and not thé gréât apes. The names above thé figurecorrespond to thé five bauplan names.

heterochronies and mutations of regulator gènesas thé engines of morphological change, andespecially of thé cranio-facial contractions thatpunctuate thé évolution of thé higher apes andhominids. From gréât apes to modem mannumerous heterochronies (hypermorphosis, hy-pomorphosis and post-displacements) hâve oc-curred during ontogeny (Chaline et al. 1998),allowing (1) thé acquisition of permanentbipedalism of Australopithecus and Homo, (2)thé increased cranial capacity of primitive formsof Homo (habilis, ergaster, rudolfensis, erectus,heidelbergensis and neanderthalensis) and (3) thédisappearance of simian characters associatedwith renewed increase in cranial capacity inHomo sapiens.

Analysis of thé neotenic mechanism in théaxolotl (Ambystoma mexicanum) provides anéloquent example linking genetic régulation(Humphrey 1967; Voss 1995) with environmen-tal conditions (Norris & Gern 1976). In thisinstance, environment, through température, isinstrumental during early larval stages in acti-vating or inhibiting gènes that control thésélection of thé most effective morphologieswithin a fairly broad spectrum of possibilities.

The gènes responsible for heterochronies, andmore particularly Hox gènes (Dollé et al. 1993;Carroll 1995; McGinnis & Kuziora 1997;Pennisi & Roush 1997; Meyer 1998), or other

regulatory gènes, may explain thé origin ofcranio-facial contractions in hominids andcertain human deformities, and thus provide abiological solution to thé human paradox of thé1 % genetic divergence versus thé large morpho-logical différence between humans and chim-panzees.

The internalist approach refiects thé pré-éminent rôle of thé ontogenetic mechanism(Chaline 1994, 1998). The next step is to examinethé climatic data that reflect thé externalconstraints on biological Systems.

Climati c patterns through time

The pattern of climates in Africa is controlledprimarily by thé position of thé equator and thétropics (climatic gradients perpendicular to théequator). Ascending air masses in thé equatorîalzone lead to cloud formation. The next factor inclimate distribution is thé présence of océans oneither side of thé continental landmass, whichprovide thé water vapour necessary for cloudformation (gradients parallel to thé equator).Only then does orography come into play bylocally altering thé gradients parallel and per-pendicular to thé equator. The present-daysituation has arisen very gradually since thétime thé South Atlantic opened up some 130 Maago and Africa wheeled round as it drifted more

188

Pongo Gorilla

i l 1« = §Sçî

5 "J

1 * i 1Ê Ei

|1i

•G

fe

Panmt

If 13l g|pE .̂hi| rj>-I\

i1

9

I is raaâdia "•.S? •""!• •* *""•(^ /r ^ ^i

iiiii

| :

1-

i ilI i i "J^~ \

si 3^* -G *s \

II 8 If^^ |

k- \^^"1 • • »

Gréât apesfundamental

ontogeny

| is s| i

i i|ii12. |IJ If• 31

la. tm

ïïMtl1

11 Sg

'rabustus

afrifanus

qfarensis

t2

•>

6î-•y

S1•b

•̂,

•4M» ç _̂neandoTal

erectus 1C M I I I !>̂ .

E^̂ ergasta-

hab&s\. J

' ̂ Homofundamental

ontogeny

Australopithecinefundamental

ontogeny

[ ,̂ spedation jk, «v«t J

R, >

Fig. 2. Phylogenetic diagram showing relationship patterns with regard to ail thé palaeontological andontogenetic data. Speciations occur by punctuation, and several lineages, such as australopithecines and earlyhuman, seem to undergo graduai evolutionary changes (continuous lines represent thé fossil record). The lastthree fundamental ontogenies of primates ('gréât apes', 'Australopithecines', and 'Homo') are shown. The 'gréâtapes' bauplan has been diversified into a gréât many species.

than 3000km northward (Scotese et aï. 1988). At15-20 Ma thé situation was already very similarto thé present-day situation and was in no wayrelated to thé formation of thé Rift Valley.

This geographical pattern of climates under-went many large changes, first, because Africa

has drifted relative to thé equator, which lay 5°further north about 10 Ma ago (Scotese et al.1988), but also because of changes in atmo-spheric circulation. In Late Miocène timetropical North Africa experienced at least fourépisodes of substantial climatic détérioration

.•;— .

CLIMATI C CHANGES AND HOMINID EVOLUTION 189

Fig. 3. Schematic and hypotheticat geographicaldistribution of climates and environments in Africabetween 23 Ma and Présent. ITCZ, In ter-TropicalConvergence Zone; IOC, Inter-Oceanic Confluence; 1,when thé sun is over thé Tropic of Cancer; 2, when it isover thé Tropic of Capricorn; I, area of continentaltrade winds; II, area of Atlantic monsoon; Ha, area ofpermanent Atlantic monsoon; Ilb, area of seasonalAtlantic monsoon; III , area of Indian Océan tradewinds; IV, area of East African monsoons. Theequator was 7° more northerly at 23 Ma, 5° at 10 Maand 2° at 6 Ma (after Scotese et ai 1988),

leading to aridity, thé last at thé end of Miocènetime (about 6-5.3 Ma) coinciding with a verydistinct cooling of thé Atlantic (Diester-Haass &Chamley 1982; Sarnthein et ai. 1982). Schema-tically, since thé work by Chudeau (1921), it hasbeen accepted that extensions of thé arid Saharazone are related in part to a reduced summer-time advance of thé front of thé Inter-TropicalConvergence Zone (ITCZ; Fig. 3), which mayeven remain blocked in thé Southern Hémi-sphère. The inter-Oceanic Confluence (N-S)coïncides to some extent with thé northern partof thé Rift Valley (IOC; Fig. 3).

From Late Pliocène time (c. 3.2 Ma) généralatmospheric circulation changed radically, forgeodynamic reasons. Events such as thé closureof thé Panama isthmus, thé opening of théBering Straits and mountain building in north-west America and Asia are thought to hâve beendécisive factors, in conjunction with variationsin planetary orbit, in triggering thé NorthernHémisphère ice âges (Ruddiman & Raymo 1988;Berger 1992; de Menocal 1995). For récenttimes, when chronology is relatively précise, itis now known that changes occur suddenly.After thé last glacial maximum, it is thought thatthé arid zone extended several hundred kilo-mètres further south on three occasions between

about 19 and 15ka I4C BP, each épisode lastingonly 500-1000 a (Durand 1995). The densehumid forest became fragmented and mountainbiotopes spread as a result of thé fall intempératures (Maley 1987).

Against this général context we will examinebelow thé impact of climate on: thé diversifica-tion of thé common ancestor, thé formation ofsubspecies and species, thé geographical distri-bution of thé différent species and thé diversityamong present-day humans. Chromosomal datamust be introduced hère that hâve not yet beenincluded in thé général scheme of hominidévolution.

Chromosomal data

A major advance in chromosome research wasthé development of R-banding and G-bandingtechniques. Opinions differ about thé rôle ofheterochromatin. Dutrillaux & Couturier (1986)considered that thé chromosome data fromprimates provide no évidence either for posi-tional effects or that altérations in heterochro-matin influence gène expression. This idea ismaterialized at thé phylogenic level by twoequally feasible dichotomie hypothèses entailingacceptance of three convergences or reversions,but a third hypothesis, where each rearrange-ment is regarded as unique, implies complexpopulational évolution (Dutrillaux & Richard,1997).

By contrast, Stanyon & Chiarelli (1982), aswell as Marks (1993), hâve argued that gènes canbe ( I ) repressed by placing them in, or near,blocks of heterochromatin, or (2) activated by ashift in their position away from heterochroma-tin régions. They set gréât store by active régionsdetected by Ag-NOR (silver nitrate) type meth-ods and this has repercussions for phylogeny.They concluded that common derived karyolo-gical features indicate that Gorilla and Panshared a common descent after thé divergenceof Homo. The distribution of heterochromatin atthé tips of thé chromosomes of gorillas andchimpanzees was confirmed by Marks (1993),suggesting a phylogenetic association betweenthose two taxa.

A trichotomic chromosomal model

The common ancestral chromosome formula ofail thèse species can be reconstructed by déduc-tion. The reconstruction is based on thé princi-ple that if two, three or four species share acommon chromosome, it is likely that somecommon ancestor transmitted it to ail of them.This is thé simplest relationship. We may also

_ . . .

190 J. CHALINEET^L.

ph

ase

3of

dis

tinct s

peci e

s1 p

olyt

ypic

bspe

cle*

l"l

Gorilla

gorilla

MM

«S H)4

i Ie

|a.

ïwg/o/jfo^~"X

13

17IS1S4

t

Pan

..„„>

ptmscus

7

rotws

afncan

ajaren.

ramu

w

•

is

ru ï"

7 modtQcd* ̂ chramosonte ,̂ v

2conmmmodifled

chromosomes

[ | pre

-chim

panzec

common ancestor 2

y

|f

common ancestor 1

7 modlfied chromosomescommon to ail Uneages2q, 3, 7, 10, 11, 17, 3»

11 imnodlfled dimnœomesoooimon to ail llneages

1,4,5,8,9,12,13,14,15,16,18 J

Homoeniitum

neandertdensts

trectus3y•3.f* ̂ Milis

3

1

Homo

siqxens

5 dutunosoRics taiheritcd fn>ni Proconsul6,19,21,22, X

AFRIC Aip«d«lion

vitat

Fig. 4. Trichotomic chromosomal model incorporating palaeontological and ontogenetic data. This modelexplains thé distribution of chromosome formulae of living species and proposes a solution for shared Pan-Homo, Gorilla—Pan apomorphic rearranged chromosomes and thé absence of Gorilla-Homo apomorphicchromosomal features. After a monotypic phase accounting for chromosomal apomorphies 2, 3, 7, 10, 11, 17 and20, thé common ancestor must hâve become polytypic and divided into three subspecies: pre-gorilla, pre-chimpanzee and pre-australopithecine. Some 5 Ma ago thé three subspecies produced thé three extant and fossilgênera: Gorilla, Pan and Australopithecus. About 2 Ma ago thé Homo lineage derived from Australopithecus togive primitive man (Homo crédits), and around 0.18 Ma ago modem humans (Homo sapiens). The figures refer tochromosomes rearrangements.

take thé view that two identical chromosomalchanges occur in thé same way on thé samechromosome in two related species derived froman ancestral form. This is statistically far lesslikely, although it is possible by convergence.

We hâve attempted to reconstruct thé chron-ology of thé formation of thé varions chromo-somes and thé successive events that marked théhistory of thé family (Chaline et al. 1991, 1996)(Fig. 4).

Among anthropoid apes, thé group with thémost primitive chromosomal formula is logically

thé one that branched off earliest in thé familyhistory. This is thé orang-utan group, whichdoes not share thé seven derived spécifie mutantchromosomes (2q, 3, 7, 10, I I , 17 and 20) of thécommon ancestor because it emigrated to Asia,where it has been eut off from thé Africanbranch for more than 10 Ma (Andrews & Cronin1982; Lipson & Pilbeam 1982; Simons 1989). Ithas been relatively stable from thé chromosomalpoint of view, only two new chromosomemutations having occurred since that time.

After thé séparation of thé orang-utan lineage,

.

CLIMATI C CHANGES AND HOMINID EVOLUTION 191

5 mutations

Ahybridization ?

pre-gonlla

Y////,2 mutations

r i i i i i ihomogeneous

population

Fig. 5. Dichotomie model explaining thé distribution of chromosomes in living species. This most parsimoniousmodel suggests that after thé monotypic phase, thé common ancestor split into a polytypic phase with twosubspecies (pre-gorilla and pre-australopithecine), and that thé populations from thé two subspecies interbred toproduce thé third, hybrid, subspecies (thé pre-chimpanzees).

thé common ancestral lineage of thé three extantgênera, chimpanzees, gorillas and humans,remained in Africa. This phase corresponds tothé 'common ancestor' implied by geneticsimilarities. That there was a common ancestoris irrefutably shown by thé formation of sevenspécifie mutant chromosomes (2q, 3, 7, 10, 11, 17and 20) and by thé rétention of 11 non-mutatedcommon chromosomes (1, 4, 5, 8, 9, 12, 13, 14,15, 16 and 18) inherited by thé three livingspecies.

The occurrence of seven characteristic mutantchromosomes found in thé gorilla, chimpanzeeand human descendants implies îhat thé com-mon ancestor was a single monotypic species,not divided into subspecies. Genetic pooling as aresuit of interbreeding produces a fairly evenspread of chromosome variety, including chro-mosomal variations that appear in isolated

individuals. This first part of thé history of thécommon ancestor corresponds to what we shallterm thé 'first homogeneous common ancestorphase' (Fig. 4).

Thereafter, thé common ancestor diversifiedand thé lineages separated, leading to thé gorilla,chimpanzee and humans. This divergence raisesa complex populational problem related to thédistribution of five new mutated chromosomes.

It is reported that chimpanzees and humansshare three mutated chromosomes (2p, 7 and 9)that gorillas do not hâve. Conversely, chimpan-zees and gorillas share two other spécifiemutated chromosomes (12 and 16) not foundin humans. In other words, chimpanzees sharethree common mutated chromosomes with hu-mans and two common mutated chromosomeswith gorillas. But gorillas share no spécialrearrangements with humans! This position is

192 J. CHALINE ET AL.

inexplicable by thé standard hypothesis of a splitinto two branches (dichotomie model) leading tothé gorilla-chimpanzee on one side and tohumans on thé other, as suggested by Marks(1993).

There is, however, a possible solution to thépuzzle: thé 'second heterogeneous commonancestor phase' (Fig. 4). To explain this majorparadox, it must be supposed that at a certaintime in their history, after thé first undifferen-tiated phase, thé ancestors of chimpanzees andgorillas were able to interbreed and acquire twonew chromosome mutations they alone pos-sessed, and not humans. This accords with thésuggestion of Smouse & Li (1987) 'that théancestors of ail th'ree taxa were still conspecific'(see also Pruvolo et al. 1991).

However, it must also be accepted that forsome time, perhaps différent from thé firstperiod, thé ancestors of chimpanzees and hu-mans were able to interbreed and to incorporatethree new chromosomal mutations into theirgenetic constitution, possessed by them aloneand not gorillas. The fact that thé ancestors ofgorillas and humans do not share exclusivemutations in their chromosomal formulae im-plies that they did not at that time hâve contactsallowing hybridization. They were geographi-cally isolated.

This model introduces australopithecines as amissing link between thé common ancestor andthé human lineage proper. They are neverconsidered by chromosome specialists for théobvious reason that, being extinct, their chro-mosomes are unknown. But as they formed anecessary intermediate stage, they must beincluded in thé model. By déduction, théchromosomal formula of thé extinct pre-humanform Ausîralopithecus can be evaluated fairlyaccurately, except for thé four chromosomesthat were rearranged after genetic isolation ofthé species (ch l / l / fusion of 2p-2q/18).

A dichotomie chromosomal model

A further possible dichotomie model suggeststhat after thé first common phase, two sub-species, pre-gorillas and pre-australopithecinesbecame geographically separated and chromo-somally differentiated (Fig. 5). Only then didtwo small populations of each subspecies meetand form thé pre-chimpanzee stock. Chimpan-zees could be hybrids of pre-gorillas and pre-au s tra 1 opit heci nés.

To account for thé séparation into distinctsubspecies, a new décisive factor must beincorporated hère, that of climate change.

Fig. 6. Distribution of thé common ancestor at théstart of thé polytypic phase. Pre-gorillas must hâveoccupied thé wet Atlantic monsoon zone (rain forest),pre-chimpanzees thé less humid Atlantic monsoonzone (open forest) and pre-australopithecines théIndian Océan monsoon zone (acacia savanna).

permanent Atlanti cmonsoon Indian

Océanmonsoon

Fig. 7. Distribution of thé common ancestor during thémaximum extension of thé polytypic phase. The pre-chimpanzee group must hâve been geographicallycentred around thé present-day Lake Victoria région.It stretched to northwest Africa across thé area northof thé Zaïre River in a forest-savanna mosaic or openwoodland environment. The pre-gorilla group musthâve been located on thé western edge of thé former,still north of thé Zaire River, in thé very wet tropicalrainforest zone. The pre-australopithecine group musthâve been located further east, in thé East African RiftValley, and further north in thé savanna beyond théopen forests inhabited by thé pre-chimpanzees.

CLIMATI C CHANGES AND HOMINID EVOLUTION 193

The internalist approach in thé climatiecontext

Climate as thé driving force of diversificationof thé common ancestor

Climatie changes (Durand 1995) would hâveinduced changes in thé environment that musthâve been instrumental in thé geographicalisolation of thé subspecies and species. Figures6 and 7 explain thé logic behind thèse changes.

The original common area of hominids incentral Africa is currently a privileged climatiezone where thé ITCZ (east-west) correspondingto thé thermal and meteorological equatorcrosses thé IOC (north-south) (Leroux 1983).Very schematically, we find thé followingfeatures.

• The permanent Atlantic monsoon domainand tropical rainforest occur in thé west;forest îs documented in thé Congo, Gabonand Cameroon basins from thé onset ofMiocène time (Boltenhagen et al. 1985).

• To thé north lies thé seasonal Atlanticmonsoon domain and thé savanna; inNigeria a seasonal subtropical climate isreported to hâve existed since thé Oligo-cène-Miocène boundary (Takashi & Jux1989).

• Further north still is thé Sahara zone;évidence of a large arid zone in NorthAfrica from mid-Miocene times is providedby rodents (Jaeger 1975) and by flora(Boureau et al. 1983).

• In thé south and east is thé domain of théIndian Océan monsoons and trade winds.Tropical rainforest, open woodlands andmountain forests are reported to hâvecoexisted since 19 Ma, and savannas andgrasslands since 15-12 Ma (Bonnefille1987; Retallack et al. 1990; Kingston etal. 1994). Local reliefs of thé Rift Valley aresuperimposed on this background patternto form an environmental mosaic that issensitive to climatie fluctuations (White1983; Pickford 1990).

Our hypothesis is that difîerentiation occurredin connection with thé environment as deter-mined by climate. Starting from a monotypiccommon stock (acquisition of seven chromoso-mal changes) (Figs 6 and 7), thé pre-gorillas splitaway in thé permanent monsoon domain northof thé Zaïre River barrier. The pre-chimpanzeesspread widely east-west, on thé northernboundary of thé permanent Atlantic monsoon

domain, ranging from Central to West Africa.The pre-australopithecines developed on théeastern margin under thé influence of thé IndianOcéan monsoons and trade winds, and spreadnorth-south along thé inter-oceanic confluencethat roughly coincides with thé eastern arm ofthé Rift Valley and westward across thé savannazone skirting thé north of thé tropical rainforestrange. Such variations in climate could hâvesegregated thé three subspecies in a yetunknown history, as fossil remains dating from12 to 5 Ma are very sparse. Few finds hâve beenmade to date: a molar from Ngorora (Kenya)older than 10 Ma, one from Lukeino (7 Ma)(Pickford 1975), a fragment of jaw with a molarfrom Lothagam, again in Kenya, dated to5.5 Ma, and thé Samburupiîhecus mandible(9.5 Ma).

A further implication concerns thé shape ofthé pelvis. As gorillas and chimpanzees arequadrupeds, and as bipedalism is thé apo-morphic feature characteristic of australopithe-cines and their human descendants, thé threecomponents of thé common ancestor musttherefore still hâve walked on ail fours.

Climate as thé driving force of diversificationof australopithecine, gorilla and chimpanzeespecies

Allopatric break-up of thé ancestral form wouldhâve allowed gorillas, australopithecines andchimpanzees to become isolated (Fig. 8), asevidenced by thé discovery of thé earliestaustralopithecines {Ardipithecus ramidus), datedto some 4.4 Ma in thé Pliocène deposits of théMiddle Awash (White et al. 1994, 1995).

Fossil finds of gorillas and chimpanzees wil lprobably be few and far between, perhapsbecause of thé lack of appropriate sedimentaryenvironments in their tropical forest habitat,where remains are more rarely preserved than inopen sedimentary environments of savannas.However, since becoming separate species,gorillas and chimpanzees hâve respectivelyacquired six autapomorphic chromosomal re-arrangements (ch 1/4/5-17/8/10/14 for gorillas;ch 4/5/9/15/17/13 for chimpanzees). It shouldalso be pointed out that thé bonobo (Panpaniscus) has autapomorphic rearrangementson chromosomes 2q and 7. Thèse autapomor-phies prove thé early isolation of gorillas andchimpanzees, and thé probably more récentséparation of thé bonobo. It is interesting tosee that thé teeth of Ardipithecus ramidus of théAwash (size, morphology, enamel thickness) and

194 J. CHALINE ET AL.

IndianOcéan

morooon

IndienOcéan

monsoon

Fig. 8. Distribution of gênera Pan, Gorilla andAustralopiîhecus between 5 and 2 Ma. The chimpanzeelived north of thé Zaire River in a forest-savannamosaic or open woodland environment. The gorillagroup must hâve been located on thé western edge ofthé former, still north of thé Zaire River, in thé verywet tropical rainforest zone. The australopithecinesgroup must hâve been located further east, in thé EastAfrican Rift Valley, and further north in thé savannabeyond thé open forests ïnhabited by thé chimpanzees,as suggested by thé discovery of Koro Toro (Chad).

even features of thé cranium hâve a strikinglychimpanzee-like morphology very similar to thébonobo, suggesting a possible filiation.

Climate and thé environmental diversificationit entails acted as controls over thé area ofdistribution of thé various species. Australo-pithecines were able to migrate from CentralAfrica northward to Ethiopia and thé Afars, andsouthward, skirting round thé Zambezi Riverbarrier, to South Africa, where their fossils werefirst unearthed (Taung, Sterkfontein, Swartk-rans) (Fig. 8).

The externalist approach implies that théearliest australopithecines could be found onlyto thé east of thé Rift Valley (Coppens 1986,1994), and not to thé west. The discovery of anearly australopithecine, named 'Abel', (Austra-lopithecus bahrelghazali), 2500km west of théRif t Valley (Chad) and thought to be 3-3,5 Maold (Brunet et al. 1995, 1996) challenges this partof thé theory. This discovery was, however,predictable by thé trichotomic theory of thécommon ancestor in thé internalist approach, asthé acquisition of bipedal gait provided thépotential to colonize ail of thé African savannafrom north to south and east to west (Fig. 8).

Fig. 9. Distribution of thé three subspecies of commonchimpanzees. The westernmost subspecies, thé blackchimpanzee (Pan troglodytes verus), inhabits Guineawhereas thé others are found north of thé Zaire orCongo River: thé typical common chimpanzee (Pantroglodytes troglodytes) in thé west in Congo, Gabon,Equatorial Guinea and Cameroon, and Schweinfùrthichimpanzee (Pan troglodytes schweinfurthi) further eastin northern Zaire (after Collet 1988).

Fig. 10. Distribution of thé three subspecies of gorillas.Gorillas are divided into western lowland gorillas(Gorilla gorilla gorilla) in Congo, Gabon, EquatorialGuinea and southern Cameroon, and eastern moun-tain gorillas (Gorilla gorilla graueri (Burundi andRwanda); thé third subspecies is thé mountain gorillaof Rwanda and Uganda (Gorilla gorilla beringei) (afterCollet 1988).

CLIMATI C CHANGES AND HOMINID EVOLUTION 195

Climate as thé driving force of diversificationof présent gorilla and chimpanzee subspecies

The répétition of Pleistocene fluctuations isthought to be responsible for thé currentdiversification of chimpanzees and gorillas, eachcharacterized by three subspecies (Figs 9 and10).

Starting from thé initial chimpanzee distribu-tion, it is thought that common chimpanzeesthen diversified into three subspecies. Onesubspecies (Pan troglodytes verus) is isolated inGuinea (Collet 1988), probably as a resuit of abreak in thé forest caused perhaps by a south-ward shift of thé Sahara désert zone inQuaternary times engendering aridity in whatis now Nigeria (Fig. 9). A second area is locatedin thé Central African Republic and northernZaïre, along thé Oubangui River, which alreadyforms a naturai barrier between thé areas ofdistribution of western gorillas and chimpanzees(Gorilla gorilla gorilla and Pan troglodytestroglodytes) and eastern subspecies (Gorillagorilla beringei and graueri as well as Pantroglodytes schweinfùrthi) and which rnust hâvebeen reinforced by thé southward extension ofarid conditions.

This mechanism of séparation of species intowestern and eastern populations may hâve beencompounded because thé gorilla and chimpan-zee populations were unable to cross thé gréâtnaturai barrier of thé marshland basin of théZaire River.

ït is impossible at présent to provide détailsand a chronology of what might hâve happened,but this mechanism is plausible. It can, further-more, be tested, because if it is correct, fossils ofchimpanzees should be found in Tertiary depos-its in areas where chimpanzees no longer occurfrom thé Ivory Coast to Nigeria.

Climate as thé driving force of diversificationof primitive and modem Homo.

Climate plays a rôle in development of théspecies Homo erectus, which survived for at leastone and a half million years and différendatedgeographically, as indicated by thé two opposingmorphological trends observed in Homo erectusof Asia and of Europe, thé latter giving rise toNeandertals. Climate also plays a further rôle indevelopment of Homo sapiens, by determiningclinal différences affecting characters related toclimate, in particular pigmentation and variousselected physiological factors (adaptations toaltitude, humidity, aridity, etc.) (Lewontin 1982;

Langaney 1988). Thèse difTeren dation s musthâve been thé same for thé various geographicalforms of Homo erectus.

Such adaptive differendation can hâve oc-curred only in thé last 180ka for Homo sapiens,which derived by cranio-facial contraction froma Homo erectus population. It is because thisdifferentiation occurred so recently on théevolutionary scale that modem humans arealone in not being subdivided into geographi-cally quantifiable and identifiable races.

Conclusion: climate as a drivin g force ofhominid évolution?

The foregoing discussion shows that climate wasinstrumental in hominid évolution, but was itreally a driving force behind it?

The externalist approach, with its claim thatclimale was responsible for thé appearance ofbipedalism, now seems to be ruled out bybiological, palaeogeographical, palaeontologicaldata and palaeoclimatic data on which it wasbased. The pattern of différent environments inrelation to thé main climatic domains (perma-nent or seasonal Atlantic monsoon and IndianOcéan monsoon and trade winds) seems to hâvebeen comparable with thé present-day situationsince about 15 Ma, i.e. well before thé formationof thé Rift Valley (8 Ma).

Biological data do, however, support théembryonic origin of cranio-facial contracdon,which determined thé increase in cramai capa-city and, through thé position of thé foramenmagnum, brought about bipedalism. Thèseprocesses are certainly partly under thé controlof so far unidentified homeobox gènes (Hoxgènes) or other regulatory gènes. The story ofhuman évolution may be more appropriatelytermed an internalist history rather than anexternalist one,

In thé internalist approach, developmentalbiology is thé major driving force behindhominid évolution but climate exerts a signifi-cant influence at crucial moments in that history:(1) in thé prior development of ecological nichesallowing thé common ancestor (polytypicalphase 2) to differentiate into three subspecificlineages foreshadowing thé three present-daygênera; (2) by facilitating relative fluctuations ofthé geographical areas of distribution of thévarious species, especially thé extension ofaustralopithecines towards thé Sahara in thénorthwest, Ethiopia in thé northeast, and intosouthern Africa; (3) by breaking up thé area ofdistribution of extant species of gorillas and

196 J. CHAUME ET AL.

chimpanzees, which led to thé three present-daysubspecies of each; (4) by determining slightadaptive différences in pigmentation, whichmust hâve been thé same for thé variousgeographical forms of Homo erectus and Homosapiens.

This work was supportée by thé UMR CNRS 5561(Biogeosciences) and by thé Laboratoire de Paléobio-diversité et Préhistoire de l'Ecole Pratique des HautesEtudes (Université de Bourgogne, Dijon). We aregrateful to two anonymous reviewers and to M. Hartfor helpful comments, to A. Bussière for thé finaldrawings, and to C. SutclifTe for help with translation.

Références

ANDREWS, P. & CRONIN, J. E. 1982. The relationshipsof Sivapithecus and Ramapithecus and thé évolutionof thé orangutan. Nature, 297, 541-546.

ANTHONY, J. 1952. La brisure de la base du crâne chezles Primates. Annales de Paléontologie, 38, 71-79.

BAILEY , W. J., FITCH, D. H. A., TAGLE, D. A.,CZELUSNIAK, J., SLIGHTOM, J. L. & GOODMAN, M.Î991. Molecular évolution of thé w-globin gènelocus: gibbon phylogeny and thé hominoid slow-down. Molecular Biology and Evolution, 8, 155-184.- , HAYASAKA , K., SKINNER, C. G., KEHOE, S.,

SIEU, L. C., SLIGHTOM, J. L. & GOODMAN, M. 1992.Reexamination of thé African hominoid trichotomywith additional séquences from thé primate /5-globingène cluster. Molecular Phylogenetics and Evolution,1, 97-135.

BERGER, A. 1992. Le climat de la Terre. De BoeckUniversité, Brussels.

BOLTENHAGEN, E., DEJAX, J. & SALARD-CHEBOLDAEFF,M. 1985. Évolution de la végétation tropicaleafricaine du Crétacé à l'actuel d'après les donnéesde la palynologie. Bulletin de la Section des Sciences,8, 165-194.

BONNEFILLE, R. 1987. Evolution des milieux tropicauxafricains depuis le début du Cénozoïque. Mémoiredes Travaux de l'Ecole Pratique des Hautes Etudes(EPHE), 17, 101-110,

BOUREAU, E., CHEBOLDAEFF-SALARD, M., KOENIGUER,J. C. & LOUVET, P. 1983. Évolution des flores et de lavégétation tertiaires en Afrique, au nord de l'équa-teur. Bothalia, 14, 355-367.

BRUNET, M., BEAUVILAFN , A., COPPENS, Y., HEINTZ, E.,MOUTAYEA , H. E. & PILBEAM, D. 1995. The firstautralopithecine 2,500 kilomètres west of thé RiftValley (Chad). Nature, 378, 273-274.

1996. Australopithecus bahrelghazali, une nouvelleespèce d'hominidé ancien de la région de Koro Toro(Tchad). Comptes Rendus de l'Académie desSciences, Série Ha, 322, 907-913.

CARROLL, S. B. 1995. Homeotic gènes and théévolution of arthropods and chordates. Nature,376, 479-485.

CHALINE , J. 1994. Une famille peu ordinaire. Le Seuil,Paris.

— 1998. Vers une approche globale de l'évolution

des Hominidés. Le Point sur-Comptes Rendus del'Académie des Sciences, Série //, 326, 307-318.

-, DAMBRICOURT MALASSÊ, A., MAGNIEZ-JANNIN,F., MARCHAND, D., DAVID , B., MILLET , J.-J. &COURANT, F. 1998. Quantification de l'évolutionmorphologique des Hominidés et hétérochronies.Comptes Rendus de l'Académie des Sciences, Série II,326, 291-298.

-, DURAND, A., MARCHAND, D., DAMBRICOURTMALASSE, A. & DESHAYES, M. J. 1996. Chromo-somes and thé origins of Apes and Australopithe-cines. Human Evolution, 11, 43-60.

-, DUTRILLAUX , B., COUTURIER, J., DURAND, A.& MARCHAND, D. 1991. Un modèle chromosomiqueet paléobiogéographique d'évolution des primatessupérieurs. Geobios, 24, 105-110.

CHIARELLI , B. 1962. Comparative morphometric ana-lysis of primate chromosomes of thé anthropoidapes and man. Caryologia, 15, 99-121.

CHUDEAU, R. 1921. Les changements de climat duSahara pendant le Quaternaire. Comptes Rendus del'Académie des Sciences, 172, 604-607.

COLLET, J. Y. 1988. Les petits des grands singes. Terresauvage. Nuit et Jour, 6-12.

COPPENS, Y. 1986. Evolution de l'homme. ComptesRendus de l'Académie des Sciences, 3(3), 227-243.

1994. Une histoire des Hominidés. Pour laScience, 201, 64-71.

DAMBRICOURT MALASSÊ, A. 1987. Ontogenèses, paléon-togenèses et phylogenèse du corps mandibulairecatarhinien. Nouvelle interprétation de la mécaniquehumanisante (théorie de la foetilisation, Bolk, 1926).Thèse de Doctorat, Muséum National d'HistoireNaturelle, Paris.

1988. Hominisation et foetalisation. ComptesRendus de l'Académie des Sciences, Série //, 307,199-204.

1993. continuity and discontinuity duringHominization. Quaternary International, 29, 86-98.

1996. Nouveau regard sur l'origine del'homme. La Recherche, 286, 46-54.

DE GROUCHY, J., TURLEAU, C., ROUBIN, M. & KLEIN,M. 1972. Évolutions caryotypiques de l'homme etdu chimpanzé. Étude comparative des topographiesde bandes après dénaturation ménagée. Annales deGénétique, 15, 79-84.

DELATTRE, A. & FENART, R. 1954. Rotation occipitalepositive et négative. Comptes Rendus de l'Académiedes Sciences, 239, 676-678.

& ——— 1956. L'hominisation du crâne est-elle terminée? Comptes Rendus de l'Académie desSciences, 243, 429-431.

& 1960. L'hominisation du crâne.CNRS, Paris.

DE MENOCAL, P. B. 1995. Plio-Pleistocene Africanclimate. Science, 270, 53-59.

DENIKER, J. 1885. Recherches anatomiques et embry-ologiques sur les singes anthropoïdes. Foetus degorille et de gibbon comparés aux foetus humains etaux anthropoïdes jeunes et adultes. Archives deZoologie expérimentale et générale, 2(3 bis), 1-265.

DESHAYES, M. J. 1986. Croissance cranio-faciale etorthodontie. Masson, Paris.

1991. Reconsidération de la croissance cranio-

CLIMATI C CHANGES AND HOMINID EVOLUTION 197

faciale au cours de l'ontogenèse et de l'évolution.Implications pour les traitements orthopédiques.Revue d'Orthopédie dentofaciale, 25, 353-365.

1997. A new ontogenetic approach to cranio-facial growth. The basis of Projet TéJécrane Inter-national. Journal of Mastication and Health Society,7(2), 59-76.

& DAMBRICOURT MALASSÉ, A. 1990. Analysedes différents types architecturaux cranio-faciauxpar l'approche ontogénique de l'hominisation.Revue de Stomatologie et de Chirurgie Maxillo-faciale, 91, 149-258.

DEVILLERS, C. & CHALINE, J. 1993. Evolution, anEvolving Theory. Springer, Berlin.

DIESTER-HAASS, L. & CHAMLEY, H. 1982. Oligocèneand post-Oligocène history of sédimentation andclimate off Northwest Africa (DSDP site 369). In:VON RAD, U., HINZ; K., SARNTHEIN, M. & SEIBOLD,E. (eds) Geology of thé Northwest African Continen-tal Margin. Springer, Berlin, 529-544.

DOLLÉ, P., DIERICH, A., LE MEUR, M., SCHIMMANG, T.,SCHUBAUR, B., CHAMBON, P. & DUBOULE, D. 1993.Disruption of thé Hoxd-13 gène induces localizedheterochrony leading to mice with neotenic limbs.Ce//, 75, 431-441.

DURAND, A. 1995. Sédiments quaternaires et change-ments climatiques au Sahel central (Niger et Tchad).Africa Geoscience Review, 2, 323-614.

DUTRILLAUX , B. & COUTURIER, J. 1986. Principes del'analyse chromosomique appliquée à la phylogénie:l'exemple des Pongidae et des Hominidae. Mamma-Ha, 50, 22-37.

— & RICHARD, F. 1997. Notre nouvel arbre defamille. L'analyse des chromosomes permet deréécrire l'histoire des primates. La Recherche, 298,54-61.

-, VIEGAS-PEQUIGNOT, E. & COUTURIER, J. 1986.Les méthodes cytogénétiques. Mammalia, 50, 11-20.

GODFREY, L. & MARKS, J. 1991. The nature and originsof primate species. Yearbook of Physical Anthropol-ogy, 34, 39-68.

GONZALEZ, I. L., SYLVESTER, J. E., SMITH, T. F.,STAMBOLIAN , D. & SCHMICKEL, R. D. 1990. Riboso-mal RNA gène séquences and hominoid phylogeny.Molecular Biology and Evolution, 7, 203-219.

GOODMAN, M., BAILEY , W. J., HAYASAKA , K., STAN-HOPE, M. J., SLIGHTOM, J. & CZELUSNIAK, J. 1994.Molecular évidence on primate phylogeny fromDNA séquences. American Journal of PhysicalAnthropology, 94, 3-24.

GOULD, S. J. 1977. Ontogeny and Phylogeny. BelknapPress of Harvard University Press, Cambridge, MA.

HAYASAKA , K., GOJORI, T. & HORAI, S. 1988.Molecular phylogeny and évolution of primatemitochondrial DNA. Molecular Biology and Evolu-tion, 5, 626-644.

HIXSON, J. E. & BROWN, W. M. 1986. A comparison ofthé small ribosomal RNA gènes from thé mitochon-drial DNA of thé gréât apes and humans: séquence,structure, évolution, and phylogenetic implications.Molecular Biology and Evolution, 3, 1-18.

HORAI, S., SATTA, Y., HAYASAKA , K. et al. 1992. Man'splace in Hominoidea revealed by mitochondrialDNA genealogy. Journal of Molecular Evolution, 35,

32-43.HUMPHREY, R. R. 1967. Albino axolotls from an

albino tiger salamander through hybridization.Journal of Heredity, 58, 95-101.

JAEGER, J. J. 1975. Les Muridae (Mammalia, Rodentia)du Pliocène et du Pléistocène du Maghreb. Origine;évolution; données biogéographiques et paléoclima-tiques. Thèse de Doctorat, Université des Sciences etTechniques du Languedoc, Montpellier.

KING, M. C. & WILSON, A. C. 1975. Evolution at twoïevels in humans and chimpanzees. Science, 188,201-203.

KINGSTON, J. D., MARINO, B. D. & HILL , A. 1994.Isotopic évidence for Neogene hominid paleoenvir-onments in thé Kenya Rift Valley. Science, 264, 955-959.

LANGANEY, A. 1988. Les Hommes. Colin, Paris.LEROUX, M. 1983. Le climat de l'Afrique tropicale.

Champion, Paris.LEWONTIN, R. 1982. La diversité des hommes. Belin,

Paris.LIPSON, S. & PILBEAM, D. 1982. Ramapithecus and

hominid évolution. Journal of Human Evolution, 11,545-548.

MALEY , J. 1987. Fragmentation de la forêt densehumide africaine et extension des biotopes mon-tagnards au Quaternaire récent: nouvelles donnéespolliniques et chronologiques. Implications paléo-climatiques et biogéographiques. Palaeoecology ofAfrica, 17, 307-334.

MARKS, J. 1985. C-band variabiiity in thé commonchimpanzee, Pan troglodytes. Journal of HumanEvolution, 14, 669-675.

— 1993. Hominoid heterochromatin: terminal C-band as a complex genetic trait linking chimpanzeeand gorilla. American Journal of Physical Anthro-pology, 90, 237-246.

MATERA, A. G. & MARKS, J. 1993. Complex rearran-gements in thé évolution of hominoid chromosomeXVII . Journal of Human Evolution, 24, 233-238.

McGiNNis, W. & KUZIORA, M. 1997. Les gènes dudéveloppement. Pour la Science, hors série, 7614,126-133.

MEYER, A. 1998. Hox gène variation and évolution.Nature, 391, 225-228.

MIYAMOTO , M. M., KOOP, B. F., SLIGHTOM, J. L.,GOODMAN, M. & TENNANT, M. R. 1988. Molecularsystematics of higher primates: genealogical rela-tions and classification. Proceedings of thé NationalAcademy of Sciences of thé USA, 85, 7627-7631.

—, SLIGHTOM, J. L. & GOODMAN, M. 1987.Phylogenetic relations of humans and African apesfrom DNA séquences in thé i/>n globin région.Science, 238, 369-373.

NORRIS, D. O. & GERN, W. A. 1976. Thyroxine-induced activation of hypothalamo-hypophysal axisin neotenic salamander larvae. Science, 194, 525.

PENNISI, E. & ROUSH, W. 1997. Developing a new viewof évolution. Science, 277, 34-37.

PERRIN-PECONTAL, P., GOUY, M., NIGON, V. M. &TRABUCHET, G. 1992. Evolution of thé primate /?-globin gène région: nucleotide séquence of thé 0-0globin intergenic région of gorilla and phylogeneticrelationships between African apes and man. Jour-

[98 J. CHALINE£r/4L.

nal of Molecular Evolution, 34, 17-30.PICKFORD, M. 1975. Late Miocène sédiments and

fossils from thé Northern Kenya Rift Valley.Nature, 256, 279-284.

- 1990. Uplif t of thé Roof of Africa and ilsbearing on thé évolution of mankind. HumanEvolution, 5, 1-20.

PRUVOLO , M. T., DlSOTELL , T., AMMARD , M. W.,BROWN, W. M. & HONEYCUTT, R. L. 1991. Resolu-tion of thé African hominoid trichotomy using amitochondrial gène séquence. Proceedings of théNational Academy of Sciences ofthe USA, 88, 1570-1574.

RAFF, R. A. 1996. The Shape of Life. University ofChicago Press, Chicago, IL.

— & KAUFMAN , T. C. 1983. Embryos, Gènes, andEvolution. MacMillan, New York.

RETALLACK , G. J., DUGAS, D. P. & BESTLAND, E. A.1990. Fossil soils and grasses of a middle Miocèneeast African grassland. Nature, 247, 1325-1328.

RUDDIMAN, W. F. & RAYMO, M. E. 1988. NorthernHémisphère climatic régimes during thé past 3 Ma,possible tectonic connections. Philosophical transac-tions of thé Royal Society of London, Séries B, 318,411^*30.

SARNTHEIN, M., THIEDE, J., PFLAUMANN, U. et al. 1982.Atmospheric and oceanic circulation patterns offNorthwest Africa during thé past 25 million years.In: VON RAD, U., HINZ, K., SARNTHEIN, M. &SEIBOLD, E. (eds) Geology of thé North West AfricanContinental Margin. Springer, Berlin, 545-604.

SCOTESE, C. R., GAHAGAN, L. M. & LARSON, R. L.1988. Plate tectonic reconstructions ofthe Cretac-eous and Cenozoic océan basins. Tectonophysics,155, 27-48.

SlBLEY , C. G., COMSTOCK, J. A. & AHLQUIST , J. E. 1990.DNA hybridization évidence of hominoid phylo-geny: a reanalysis of thé data. Journal of MolecularEvolution, 30, 202-236.

SIMONS, E. L. 1989. Human origins. Science, 245,1343-1350.

SMOUSE, P. E. & Li, W. H. 1987. Likelihood analysis ofmitochondrial restriction-cleavage patterns for théhuman-chimpanzee-gorilla trichotomy. Evolution,41, 1162-1176.

STANYON, R. & CHIARELLI , B. 1981. The chromosomesof man and apes: évolution and comparison.Bionature, 1, 11-27.

— & — — 1982. Phylogeny of thé Hommoidea:thé chromosome évidence. Journal of Human Evolu-tion, 11, 493-504.

— & — - 1983. Chromosomal phylogeny ingréât apes and humans. Antropologia Contempor-anea, 6, 157-159.

TAKASHI , K. & Jux, U. 1989. Palynology of MiddleTertiary lacustrine deposits from thé Jos Plateau.Bulletin Nagasaki University of Natural Sciences, 29,181-367.

UEDA, S., WATANABE, Y., SAITOU, N. et al. 1989.Nucleotide séquences of immunoglobulin-epsilonpseudogenes in man and apes and their phylogeneticrelationships. Journal of Molecular Biology, 205, 85-90.

Voss, S. R. 1995. Genetic basis of paedomorphosis inthé axolotl, Ambystoma mexicanum: a test of thésingle-gène hypothesis. Journal of Heredity, 86, 441-447.

WHITE, F. 1983. The Végétation of Africa. UNESCO,Paris.

WHITE, T. D., SUWA, G. & ASFAW, B. 1994. Australo-pithecus ramidus, a new species of early hominidfrom Aramis, Ethiopia. Nature, 371, 306-312.

, - -— & 1995. Australopithecusramidus, a new species of early hominid fromAramis, Ethiopia. Corrigendum. Nature, 375, 88.

YUNISH, J. J. & PRAKASH, O. 1982. The origin of man: achromosomal pictorial legacy. Science, 215, 1525-1530.