Osborn 1902

The Law of Adaptive Radiation Henry Fairfield Osborn The American Naturalist, Vol. 36, No. 425. (May, 1902), pp. 353-363. Stable URL: http://links.jstor.org/sici?sici=0003-0147%28190205%2936%3A425%3C353%3ATLOAR%3E2.0.CO%3B2-1 The American Naturalist is currently published by The University of Chicago Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/ucpress.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Mon Aug 20 15:04:28 2007

AMERICAN NATURALIST

T H E LAW O F ADAPTIVI;

R-IDIATION.

OSE of the essential features of divergent evolution as conceived in the branching system successively clevelopcd by Lamarck, Darwin, Husley, and Cope has been termed by the w r ~ t e r<
THE 1

354

I A

U

RI S .

[VOL. S S S V T

principles of hon~oplasy,parallelism, and convergence, which, as s h o ~ v n in t h e last number of t h e .i;rtz~i-lzlisf, a r e by no means synonymous terms or identical processes. T h e altogether similar la\\- of Zo~rrZnclajti.iv rrril7rizfz'oi~01- incipient divergence on a s ~ n a l l e rscale in a single locality may 11ow be Inore clearly developed. This idea of radiation beco~llesa means of interpretation, and a way of imagining t h e relations of extinct and living faunze. A s percei\.ecl by Cope, it applies both on a small and on a vast scale.

p p. 3 0 - 3 3 ) t h e ancient I n t h e Rise of t h e ;Ilammalia " (193, Mesozoic (LIeseutheria) and modern Cenozoic (Ceneutheria) differentiation of the placentals (Eutheria) was spolten of as follows : T h e I'uerco is e s ~ e n t i a l l yan archaic fauna, a n d is to h e regarded a s t h e climax of t h e first periocl of placental differentiation: a culmination of t h c first attempts of natilre to establish insectivorous~ carnivorous, and herbivorous groups. T h e s e attempts began in the Cretaceous, and some of the types thus produced cliecl o u t in the Puerco, sonle in tlle IYasatch a ~ i d nridger 1 only a fexv flesh-enters surl-ive to the LZiocene. I t is no st important to g r a s p clearly the idea of this ~ ~ c l i ( ~ f i i ~1 ~i i ~i zi f~i i ~ f ill i O iall ~ directions of this old Puerco fauna, resulting in fornls like the modern insecti\.ores, rodents, bears, dogs a n d cats, monkeys, sloths, hunoclont and selenodoi~t ungulates, a n d lophodont ung~1la:es. T h i s \\-as a n independent radiation of placentals, like the A~istralianradintion of marsupials. S o m e of the least speciallzeti spurs of tliis rndiation appear t o h a v e survivecl and become the centers of t h e second o r niici-Tertiar!. radiation. from which our moclerii fa~una has el.oli-ecl. Yet we have not in a single case succeeded in tl.acing tlie clil-ect connection.' T o sunl up, we find on tlie North ; i ~ n e r i c a ncoiltinent evidence of the rise anti decline a n d disappearance of monotremes a n d mars~ipials. and two g r e a t periods of placental racliation, t h e n/ccit.i~frizrl'iitfioii begitlning in the LIesozoic, reaching a cliinas in t h e Puerco a1lc1 unknolr-n post-l'iierco, ancl sending ith spurs into t h e higher Tertiar!., and t!le //ioiJt.i.n r.nil'i'izti'oi~reacliing its clinlax in the Lliocene, ancl sendillg d o ~ v nto 11s our esisting types. 1

This statz~uent11x5 been indclified by subsequent discovery.

No.43 j.1

I n the

L A li'

O F ADAPTIL'E

Origin of t h e Mammals developed as follo~vs: ii

RADfrl TIOA'L: "

(99,

355

p. 92) tlae idea was

T o guicle our speculation in tlie ~ ~ n l c n o ~pre-Tertiary vn periocl, we m a j gather certain positive prillciples from tlie 1;non.n evolution of the Tertiar! hlammalia. First, n e Icnon. that i z i i f l j f i z ~ c~ ~iaiZthfi'oiz, characteristic of all vertel~rates,a n d beautifully illustrated anlollg lieptilia, is in a very lligll clegree distinctive of 5Iammalia. because of their superior plasticity. Tile fbcnZ-types, or nlost primitive forms of the racliations, I-IT-, V J L , ) . ~ ~.f o r oiit iz i7,ot.o / / s ,fiut-i/~.$ , cerfiziii& sii~irZZ, t ~ ~ i ~ ~ ~ ~ ~Ls .tZi ~ Zz '~i Lzi i,l /L,~~, ,ei:,o).o/r.s I t is ilotelvortliy t h a t in the evolution of each radiation, s o far a s lve Itnoiv a t present, land types and organs a r e invariably primitive, a n d water types a n d organs are secondary, exactly as we find a m o n g the Iieptilia. I11 fact, we 13ave n o t foulld a siilgle iilstance in \vliich a mammal or reptile series is I<no~vnt o b e transformed from a water into a land t y p e : it is always tlie reverse. Tliere is certainly n o evidence for a cetoid ( X l l ~ r e c h t ) stein of the ~ T a m m a l s . Again, it is obvious that ileitl~ercarilivorous nor herbivorous types with liigllly specialized or reducecl teeth anti feet c a n lje s o central a s iilsectivorous a n d omnivorous types. I n fact, tile Iilsectivores a m o n g Place~ltals.ant1 O p o s s u ~ n sanlong Rlarsupials, are tile onl!- aninlals which have preberved the dental protot!-pe close to that of tlie P r o m a n i n ~ a l .

T h e radiations spoken of in this essay were :

I11 a subsequent paper, " T h e Geological ant1 Faunal Relntions of Europe and America during t h e Tertiary Periotl, arid t h e Theory of t h e Successive Invasions of a n African F a u n a " (loo), t h e subject was further develol~erlas follo~vs:

I q o v it is a well-knoivn principle of zoological evolution that an isolated region, if large a n d sufficiently varied in its to!~ograpliy, soil. climate, anti vegetation, will give rise to a diversified fauna accordin,q to the law of adaptive radiation from pl.imitive and central t rel="nofollow">-pes. Branches ill spring off in all directions to take aclva~ltaqe of every possil~lc opportunity of securinq food. T h e inodifications ~vliicll animals undergo in this adaptive radiation are iargeiy of m e c h a ~ l i c a lnature. they are linliteti in tiuml~es

356

THE A,lfEIIICA ,Y 11.-ATURALIST.

[ V O I ~XXXV , I,

ancl kind by hereditary, stirp, or germinal influences, and thus result it1 tlie inciependent erolution of similar types in widely separated regiolis under tlie Zn.ii~c~'$(zi.alZcli.r//i oi* iloi/cc~$ing~.~

11. ADAPTIVE R A D I A ~ I of O NORDILRS A S D F U I I L I E SA s ITG O N GEOGRAPIIICIL DISTRICU~IOK.

GEAR-

This lalv causes the inclepe~ide~it origin not otily of similar genera but of siniilar faillilies and even of similar orders. S a t u r e tlius repeats herself JVlien 011 a vast scale, but the si~ililarity is never colnplete or exact. migrations are favored by over-population or geographical changes, a new and severe test of fitness arises by the miilgling a11d competitioa of the parallel types. Under tile operation of tliese laws a 111ost interesti~iggeneralization or hy1)othesis can b e made a s to tlie tliree [zoological] realms: geographical

FIG.I . - Orders of manlmal? placed in their llypotlleticnl chief Centers of aclalitive r.idiatiu11 daring the Tertiary period.

(From Orborn.)

isolation has been so co~ltilluousand prolo~lgedthat great orders of Inammals liavt: heen evolvecl . . . io each. T h u s Ai-cta~en, coiltai~iing tlie l~roaclestand most highly- diversified lalid area, appears hypothetically as tile center in which fourteen primitive and specialized orders radiated fro111 each other. I n the southern portion of .\i.ogenat least four orders sprang from pri~ilitive meinbers of tlie above orders, and the Hystricomorpli rodents enjoyed their chief radiation. In .\fqqeiz two orclers were cut off by the s e a ; one of thein a rapidly declining type, tlie hIo~iotrernes,the other, the Alarsupials, enjoy-iiig a very highly diversified radiation. This At this time the distinction between homoplasp and parallelism was not appreciated by tile writer.

h o. 42 j.]

L A l i * OF / l I j ~ l l ' T lP-E f?AlllA4TIOL1.

357

hypothesis is expressed in Fig. I I1 [Fig. I 1, T h e other orders of malnmals, the S i r c n i ; ~(111-ol,al,l!- a l~rancll of the hoofed t r i l ~ c ) ,took tlie ri\-e:-s ;11itl coasts of :Imerica, Lul-ope, a n d proi~;lbiy X f 1 . i ~a~s ~their rac!inting center. while the Cctacen occupied t h e l'ourtli o r oceanic realm. '\Ye mean to c s p ~ c s 11y i this li!-pothesiz that I < i - . ~ ~ , \ i i[Fiq. 21 .iijei.i. / h L , ?/>lzi;~ ct;~td/.s lz
(v-

P I G ,~ , - I I i v i s i c ~ n o f t1.w \\,>11.~1 in.<, tI.11ec !ealn,,\
iF1.0111( ) i b o r ~ ii .

means tlie e s c l u s i ~ careas o l distribution, for during the periocls of lnntl contact certain n1enl11e1.s of t ! ~ e s e ortiers found tlieir n-a!- into adjace~l! realms. C n c l ~realm, tlierefi~re,c o ~ l t a i ~ its l s pure autocl~tlionoustypes antl its migrant or derivecl types. IZrc;ic~si,o n tlic other h a n d , m a y b e dibtinguislied from realu~is as geographical aiid 7oiilogical areas. vrllich lial-c been iiolated from encli other for shorter pel.ioda, either 11)- climntic 1xi1.riers, a s in the case o i tlie Arctic or cil.cmmpolar r e g i o ~ i ,or l ~ ygreat phpsical barriers, s u c l ~as masses of water ancl of clesert sailclx. I n certain cases these regions. sucli as Africa, appear to haye l ~ e e nso largi., tlisti~lctantl isolateci, a s to !ial-e l ~ e c o m eirnporta~ltcenters of t h c radiation of certain oriicilr o i inammals and almost attain the rank of realms, but reyions in general are chiefly a ~ i tpi e r m a n e ~ i t distinquished l 11y the i z i i i 7 j i i ; ' c izz(ii1riiiiii (1

j~f'7//1ilici q j . ;//lr;;//1/(7l,i.

I n this paper Africa ~ v a sti-eatecl hypothetically as a great center of intlepel~tletlte\~olutiollanti as t h e s o ~ 1 l . cof~ succesui\.e migrations of animals. T h i s hypothesis has recent~ly nol.th~~arc1 been col~firnlecl by remar1;able pnleontolo~ical discoveries in northern Africa. T h i s ndtls t o t h e above list of five radiations a sisth, namely

358 1-1.

THE A,IIER/CALV *\-A TTCTRALIST.

[VOL. X X X V I .

ft-icnliXnn'intla/t o!f PZ/rci,r~tnZs, chiefly Plol~osciclea; Hyracoidea; the families Antelopidie, Gi~affida,Hippopotamidre, etc.

ylc. 3 , - ~ ~ \ t ~of ~. i 2 tan l -t c t i c a~a, 11~~10tlietic.il ~ center of T e l t i n l y ad.il~tiveiaiiintioi~,1)). A vatiorl to the j u j u sounding line, s h ~ \ing \ uld colitil~cntallilies. (I'roin O s b o ~ n . )

As seen in operation among t h e ungulates the competition ancl range for food originate5 t h e lengthening of limb fro111 slo~v-footeclinto cursorinl types, aucl t h e lengthening of teeth from short-crowned (brachyodont) into long-croivnecl (hypsodont) t ~ p e s and , frecjuently t h e lengthening of skull from blachycephalic into clolichocephalic types (Osborn, '02a). T h e especial application t o paleontology t o b e noted here is that as these types may have lived together or in proximity and rcsortecl t o t h e same water courses for drink, their foisilized

remains a r e often fo~intltogether. STet if we csalllille analogous types living to-clay, we see that they do not frequent exactly tlie same feeding ranges nor (lo they subsist upon esactlp t h e same food ; they t h u s do not compete. Goocl illusi ~ in t h e distributrations of this iociri rrdivlz~c i t t ~ i ~ t ~ail ~e oseen tion in Africa of t h e closely allied squaic-llppecl lihzi~uctios s i ~ i ~ iwith l s 11) psocloilt teeth, ~ ~ h i c lives l l upon grasses, and t h e nnol e pointetl-lipped 12, bl'cornis with hrachyoclont teeth, ~vllich lives mainly upon shlubs. Not11 species belong to t h e same phylum of rhinoceroses. An1ong fos5il types closely 1-elated to t h c above we obsel-I e siiniiar cases, such ns t h e coexistence in t h e lorver Pleistocene of t h e hypsodont R.ir7~tipuztat~s (~voolly rhinoceros) a n d t h e brachyodont I?. ~ L ~ I I Z ~ ( P C ~ I ~ S . If carried farther than in t h e above instances, we may b e justified in placing these I'ui.izL /~(t'irpiiw~.trii'ic~tioirs in separate subfnmilies, l ~ e c a u s ein many cases they give sise to distinct a n d long-persistent collateral phyla. Esnmples of this 1;intl a r e numerous among t h e ancient I'erissodactyla or ungulates relatctl to t h e horses, tapirs, a n d rhinoceroses, as seen below TI. S I . I C \ I ) I ~ , ~ L - F ~ O ~ ~ ~ , I J , 1,ltill ~ - L 1 l I l ~ F , l l , CIX~URIAI.

Ftr nzi/iL.s. I

.YII/Y~I~JII/~< .s.

. 1 ' ~ i l z o t l ~ e r i i ~ l ~ I'alzotlierii~ix B.l (\Iicltlle Eocene to I,on-er

2.

Tita~iotlieriid~

.;

I1yr;~cotlolititl:c

Oiigo~ene) 1'nln:osyopin;e 1;. (llitltlle Eocene to l-l'l'er l~ocene) 1Iysa~liyin:e fi. (.\Iitlt!le to L'ppel. Eocene) L o p l i i o t l a ~ i t i ~ iI x ?.(\Ti& tlle to I:pl>er E o c e i ~ e )

.+. 1.opiiiotlontitiz

Sll(L1iU

~IIIYI,I~;L.

//i',.r.

l'alaplotlie~iinx (\litltlle Eocene to 1,on e r Oligocene) 1 c i m a t o t o t l i e r i i ~ ~ , i11.l : (\litltllc Eocene t o 1.1'per Icocene) 'I'riplol~otlins /<. (\lidtile to Yppel. 1:ocene) I Ielaletinre 6. (E:oceni.) C'olodoutinn: 6.(Oligocene) 7 .

I t xvill be observed a t ollcc that there is n o inherent correlation b e t \ ~ e e niii-~r~hj~oi?c~~iti.siii z ~ i i t lbi~~r1-hjpo~(j1, or hj;hiofJoiitis~i~ ailti r J ~ I l ' ~ / i c ~ ~ 01t < lcloi~gation ', of t h e feet, a s we might have anticipated, although hypsodontism is grad~iallydeveloped in 1

B.= l,rncIij otlont : 11. = I ~ y p ~ o t l o n t

360

THE A,lIBXZC.-I*17

A\;I (C7A'ALZ.\

T.

[I-oL. SSS171

long-footed series because subsisteilce upon grasses is associated with such collditiolls of life as a r e affortlecl by estensive open plains, long ranges for food, and rapid flight from enemies. Again, as shon-n else\vhe~-e, clolichocephaly anti clolichopody, brachycephaly and b r a c l ~ y ~ o t layl e frequently but not invariably correlated. ( S e e Osboril, '02a ) T h e value of this law oi lorrri ndafti;~t7.irtr'l'iztziirris especially great as a nleaiis of interpi-etntion of t h e frequent contenipo~,l neous existei~ceo r association of more primitive (hi-achyotlont) wit11 more specialized (hypsodont) types. A m o n g Tertiary inachxroclont cats it is seen in t h e contemporary long-limbetl Drrizc-ti.s ant1 short-limbed E-(cp/ip/loi~i.lc~. 111 fact, tlle association has been so often observed that if we find one l,hylun~,we may allnost anticipate or predict t h e discovery of t h e other T h e law is made more cle,~rby referring to t h e above table and t h e following explanntion of i t : ( I ) A s cornpalet1 with t h e Pnlxotheriinx, t h e Pa1aplotheriin:e a r e so long-footed that I-Iuxley believed that t h e y gave rlse to t h e llolse, and h e actui~i~iicis as t h e ancestor of t h e horse ally placed Ptrloj/c~i'/i~~i~iiii series. ( 2 ) T h e Telmntotheriinz are large animals also tlis tinguished by very long, s l e ~ ~ t l elimbs r ; they intlepentlently acquired horns, and t h e species T. c o i . ~ ~ n t ~ nras c i i ~ believed by Osborn to be ancestral to t h e titanothel-es ; but as a matter of fact, as sllorvn by Earle and I-Iatcher, this horn is a pure cast. of parallelis~l~, since t h e titanotheres probably sprang from t h e short-skulled and I-elatively stout bpecies, i ' t r / @ o s j ~ o jiilnrlfir1~ C L Y O S , a member of t h e Pcr1n-osyopinc~ ( 3 ) T h e Triplopotlin,~ present t h e most estrenie instance of light-limbed tlevelopillent anlong t h e perissodactyl ungulates, since t h e liilibs hax-c t h e ~ ~ r o p o i - t i o nof s sonle of t h e nlost slender and swiftfooted ruminant mammals, although these animals a r e fountl in t h e same beds \I-ith t h e corl-elatetl subfamily H y r a c h y i n x . (4) Anlong t h e L,ophiotlontitl~c,t h c I-Ielaletinre bring out with additional force t h e 131-inciple that this elongation of limb occurrecl at a very early geological periotl ; because t h c first linown member of this phylum appears nray clown in t h e 1,owelEocene, \Yasatch, 01- Suessonien betls, in t h e genus H ~ f o t J o i i C o p q a remarkably light-linibecl form, ant1 it runs right throug!~,

?JLOJ/

so far as Tve know, into t h e Oligoceile genus CoiorJori, t h e threetoed type with extraordinarilq- elongate digits. S o far as we know, this light-limbed series is fouiltl both in America and Europej while t h e heavy-limbed 1,ophiodontina: a r e iouild only in Europe. Since t h e a l ~ o v e was written t h e titanotheres h a ~ r ebeen nlore carefully examined by t h e writer (Osborn, 'Qzbj, and, like t h e rhinoceroses, they are found t o subdivide into four contemporaneous phyla clisting~~ishedchiefly by clolichocephaly ailcl brachycephaly and by relatively long ant1 short limbs, t h u s affording another conspicuous illustration of this itr7sr cf Local adr$fi-;,i. ~,ndi(rfior~.

I n t h e careful consideratioil of adaptive ratliatioii from certain stem types is to be fount1 t h e t r u e significance of Cuvier's LITZU of C O I ~ ~ F L ' ~ Z ~ Zas O I I moclified by t h e - to him - unl;no~vii principle of evolution. Referring to t h e diagrams, 1;ig. 4, two inlportant princil)les are brought out : First, practically all t h e adaptations li110~v11among mammals have arisen by combinations of divergence independently pursued in t h e linlbs ancl t e e t h ; for example, a n 11erbi1-orous tooth t y p e may combine with a terrestrial, arboreal, or volant limb type, accortlinc as t h e search for plant food is on t h e earth, ill t h e trees, or in , t h e air. Xlthougli every imagina1)le combination ( r . ~ .acluatic limbs, m y r m e c o p h a g o ~ ~dentition) s cannot be realized, yet these ~ir/ combinations have been multiplied almo5t ttd i i i / i ~ ~ i t ~ancl constitute t h e fatal defect of Cuvicr's law as h e conceivctl it. A s tested by a single case, t h e Eocene monkeys of t h e family Notharctitlre acquired teeth exactly l~omoplasticwith those of Eocene horses, but t h e former were provicletl with arboreal, t h e latter \vith terrestrial, limb types. Second, correlation of lill-tb aild tooth structure in a given group is further contlitioned by t h e particular c o i ~ ~ l ~ i n a t ai o~nl dclegree of specialization of linlbs ant1 teeth which t h e radiatioi~ originates ~ v i t h . lior example, t h e primitive placentals coml~inetl t r i t u b e r c ~ ~ l a r insectivorous teeth \\-it11 a generalizetl or probably terrestrial

type of feet ; t h e Australia11 marsupial radiation, on t h e other hand, began \\.it11 a dental type siil~ilarto that of t h e p l a c e i ~ t ~ ~ l s , LIMBS AND FEET \~OL.IST

Short-linil~etl,pla~ltigl-,itle, ~11113LTl.~A~l~(~)l<\. pe11r;~tlacr jl, unyuicu0,; late s t e i n TLI
\

TEETH O>IX~~(~~
( Glass

ITell~

but, as Dollo ant1 Uenslcy har-c she\\-n, t h e foot tjrpc was of highly specializctl arbol-cal charactel.. T h e s e tn7o fundamental escel)tions makc clear n-lly it is impossible, as inany writers lnr-c observetl, t o reconstruct an

entire animal froill either a c l a \ ~ o- r a tooth. T h u s , while the la\\, of c o r r e l a t ~ o nis no less dominant than Cuvier supposed, only a vestige is left of t h e nlode of a r c h e t y ~ ~ aoperation l of t h e law a s he conceived it. I t may he now restated a s follows : F e e t (correlated chiefly with lilnb and hotly stlucture) a n d teeth (correlated chiefly with sltull and necli structulc) diverge independently in adaptation respectively to securing and eating food under different conditions. E a c h evolves tlirectly for its o\iTrn mechanical functions or purposes, yet in such a manncr that each subservcs t h e other. Con-elation is therefore not morpl~ological,as Cuvier sup~ o s e c l ,but physiological, function aln.ays preceding structure. I t beconles closest nhel-e teeth ancl feet combine in t h e same functioil as in t h e prehensile cznines and c l a ~ v sof t h e F e l i d z , and 1110st diverse where t h e functions a r e 111ost diverse, as in t h e teeth and paddles of t h e Pinnipedia.

BIBLIOGRAPHY. '68 COPE, E. D . O n t h e Origin of Genera. P?.oc. Ainii. ,l;ri. Sci. ( O c t o l ~ e r ,I 86s). P p , 242 i 9 fs t y . '93 C)SI:OKS, TIEXI{\- 1.; 'The Rise of t h e i\Iammalia i n North r\merica. c I .o A i 1701. xlii ( I S93), pp. I 87-237. '99 OSI:IIRS. HCSRYF. Tile Origin of 1Iammals. Aiiic/-.Joiii-ic Sci. 1-01.vii (Febl.uar!-, ~ S g i ) ) .1 ' ~ .92-96, '00 OSL:OKS,HESIIYF. 'I'he Geological ant1 Faunal 1:elations of Europe ant1 Amel-ica (luring t h e 'Sel-tinry Period. and the 'Tlleory of the Successive Invasions of an ;Sfl.ican F a u n a . .Yticiii~>,S.S. \-ol. xi, No. 276 (April 13. r g o o ) . p p . j61-574. '02a 0s1:o11s.HESIII-1'. I~olichocephnlyant1 I~rach!~cephalyin the L o v e r AIammals. B2tii. Aiil~.i-..liits. r l i z f . Z1L.i'. l 7 o l . xvi. art. vii, 1211. 77-39 '02b 0sr:oils. F I i - s i { ~F. Tiic Fom. l ' l ~ j l a of Oligocene Titanotlieres, Z A I l f i t 1-01.x i , art. viii, 1111. 91-109.