Ac ta P alaeontologica P ol on ica--- --Vol. 28, No. 1-2 pp. 75-92 Warszawa, 1983

Se cond S ymposium on Mes ozoic T errestial Ecosystem s, Jad wisin, 1981

ANDRZEJ ELZANOWSKI

BIRDS IN CRETACEOUS ECOSYST EMS

ELZANOWSKI , A. : Birds in Cretaceou s Ec osy stems. Acta Palaeont. Polonica,28, )- 2, 75-92, 1983. .

At lea st three ecological typ es a re di sti ngu ished among the known Cret ac eo usbirds: p iscivores , shore birds and terrestri al bi rds. Striking r a ri ty of ter restria lbi rd s is considered as a specia l case of the rarity of smaller (1- 10 k g) am ongth e medium -sized vertebrate sp ec imens in the Cretaceous r ecord. T h is is proba bl yca us ed, inter a lia, by rept il ian sca ve ngers whi ch swallow fo od item s as largeas po ss i ble, and de ca lcify bo nes co mpletely or nearly so. Theref ore, the small estch ance for p rese rvation w ould be for those an imals which a re small en oughto be swall owed w hole and large enoug h to be well de t ect a ble an d/or accessible.T he g rea t ab unda nce of la rge and f lightles s p iscivorous birds (Hesperornit hes )in the war m seas of Wester n I n terior contrasts with the lack of comparablefo r m s in the Cenozoic warm seas. The extinctio n of toothed bi rds m ay havebeen ca used by the explosive r adiat ion of acanthopterygian fis hes.

K e y w 0 r d s : birds, Cretac eou s , extinction, Hesp erornithes , Ich thy orn ithes ,m arine communities, Mesozoic, North Am eric a, p al aeoecology, sca venging, ta­phonom y, vertebrates.

A ndrzej Eltanowsk i. Zaklad Paleobi o l ogtt, P olska A k ade mia Nauk , 02-089 W ar­szawa, al. Z wtr ki i W igurll 93, Pol and: R ec ei ved: December 1981.

INTRODUCTION

Twenty three valid genera , including forty one species, have beenestablished for the Cr etaceous avi an fossils. Undescribed or unnamedmater ial of which I am aware, contains about 'seven genera with sixteenspec ies , including new he sperornithiforms announced by Martin (1980)and the Enant iornithes described by Walker (1981). Thus the approximatetotal nu mber of fossil genera ' k nown from Cretaceous deposits would bethir ty.

Within the Cr etaceous, the oldest avian fossils are isolated featherscoming from two r emote areas of the world: Australia (Victoria) andLeb anon. The four Australian fe athers (Rich 1976) come fr om the lacust­rine claystones of Korumburra Group, now considered to be of Kimmer­idgian to Alb ian (or later) da te (Molnar 1980). The Lebanese fea ther s

76 A N D R Z E J E LZANOWS K I

(sever al fragments) were fou nd in the amber of Hauterivian age (Sch lee1973). These finds indicate a worldwide distribution of birds by the earlyCretaceous (Talent et al. 1966 ; Brodkorb 1971) unless feathers occurredin some ar chosaurs, close to the avian ancestry, as has been suggested(J ens en 1969).

Preserved in the Cretace ous beds are also footpr ints of bir ds. Theoldest of them, named Aquatilav ipes swiboldi, are of Aptian age andcome from the deltaic Ge thing Formation, British Columbia, Canada(Currie 1982); they resemble the footprints of modern shorebirds, althoughthis does not substantiate an ordinal allocation to the Charadriiforsnes asthe describer did. Much younger tracks, also referred tentatively to Cha­radriiformes, we re reported from the Maestrichtian Coli Toro F ormation,Patagonia (Casamiquela 1980). Qu ite different avian footprints, having theimpression of hallux and named Ignotornis mcconnelli, come fr om theCenomanian sea sho reline dep osits of the Dakota Sandstone, Colorado(Mehl 1931). .

Besides feathers and footprints, the lower Cretaceous beds yie ldedonly three gene r a of birds. Oldest of them, GaHornis and W y leyia, donot reveal any trait that would be indicative of their habitats, Gallornisstraeleni Lambrecht, coming from undetermined Neocomian beds ofFrance (Department Yonne, Auxerre), is based on the proximal end offemur, which does not allow for any reliable assignement.Wyleyia val­densis was ba sed on a single humerus from Weald Clay of England (Har ­rison and Walker 1973). Despite Brodkorb's (1978) statement tha t it is"almo st certainly a reptilian humerus", the avian nature of W yl eyia iscor r oborated by comparison with Mongolian birds (Elzanowski 1981). Thethird early Cretaceous genus is Enaliorni s, with two species describedfrom the Upper Greensand of Albian age, England; they were foot-pro­pelled diver s, probably the oldest known hesperornithiforms (Martin andTate 1976). The Albian occurrence of such a specialized form shows thatbirds had strongly diversif ied in the early Cretaceous. Moreover, theAlbian marine limestones of the Toolebuc Formation, Queensland, Aus­tralia, yielded a small (length of centrum: 3.5 mm) vertebra which showsgeneral similarities to the cervical ver tebrae of Ichthyornis (pers. obs.based on the photograps furnished by R. Molnar, Queensland Museum)and thus appears to be avian.

Late Cretaceous birds come mostly from marine or non-marine coastaldeposits. Among a dozen of pre-Maestrichtian formations the most prolifichave been those of the Western Interior basin in North America (seebelow). Also the Maestrichtian birds are known mostly from North Ame­rica : from"coastal pl ain deposits of the formations Lance, Wyoming, andHell Creek, Montana ; and sea shor e deposits of the formations Navesinkand Hornerstown, both New Jersey. In Eu rope, the only Maestrichtianbird fossils come from the coastal plain deposits of the Hateg (Hatszeg)

BIRDS I N CRETACEOUS ECOSYS T E M S

Ta b l e

Cret a ceous genera of bi rd s and their stra tigra ph ic ranges

77

<: :I: '" :>- :>- o >-3 o Ul o [;i= Ul

'" '" '" '0 t-' CD C 0 '" ~IT

t-' C " r-t- rr ~ " ~ ~ CD '" Ul

'" rt " .... .... 0 0 .... rr '0 III .o C~ CD CD '" '" 3 ~ '" 0 '" rr CD 0 tr

'" " 3 ~ ~ '" .... o ~ ~ " " 0 o.... .... .... ~ '" .... .... .... .... o, " t-'

'" -c '" .... ~ III '" '" oC) CD '"~

.... ;J '" ~ ~ ~ ;:r" III.... '" ~ rr- CD III

~ ~ .... ~

'"c

~ III

I ••• Wy Zey i a ---

•• 1 • GobipteryJ: GOBIPTERYGI FORME S

•• ENANTI ORNI THES

Iaht hyorni a ICHT HYORNI THES---------

•- Apat ol'ni s ?- Enal.ior nie

Heepero vnieHESPERORNITHES

•- Baptornis

•• Neoqaeornie

~ Torotix ?

- ----- - - - -- Lonahodytes

· - Cimolopte rux

- Pa Li.nt.ro pus CIIARADRI IFORMES- ce ramor ni e

~ Pal.aeotn-i. nqa

- - - - - -----~ Te l matormi-e ?

~ EZopt el'yx?PELECANIFORMES

~ Gracu l.auue

~ AZexol'nis ALEXORNIT IlI FORME S-- ---- - -- -

•• • •• 1.1 Gal lorni. oNO EVIDENCE

~ Iaol'nis* FOR SYSTEMATI C

~ Parusaani ornis PLACEMENT,

* Olson (1974) sugg ested that Laorn is belongs wit h the suborder Lari of theCh arad riiformes.

78 A ND R ZEJ E L Z A N O W S K I

Basin, Romania (Grigorescu and Kessler 1980). True continental depositswith fos sil birds are extremely scarce and r epresented only by the San­ton ian or Campanian Bar un Goyot Formation, Mongolia (Gradzinski andJer zy k iewicz 1974), and th e ?Maestr ich ti an Lecho Formation, Argentina(Bonapar te and P owell 1980).

Recen t discoveries brought about a profound change of views onthe Cr etaceous avifauna. Until now, only the toothed birds h ave beenallowed, because of compelling evidence, to be primitive and not typicallyavia n in all characters. Other avian fossils (as W yleyi a or Gobipteryx)have been eith er r ej ected as reptilian (Brodkorb 1978), but without sh ow­ing similarities to any known group of reptiles, or included into moder ntaxa on the basis of more or less remote resemblances. Fortunately it hasbeen r ecently sho~n that the affinities of Mesozoic (and ev en earlyTertiary) birds are impossible to determine by small fragments (Olson andFeduccia 1980a) and, as in, the case of leaves of Cretaceous plants, thepl acement of many fragmentary fossils of Cretaceous birds into moderntaxa is unwarranted. A sceptical look at the earliest record leaves theCharadriiformes as the only recent group that is well evidenced in theCretaceous (Feduccia 1980). Alth ough this picture is certainly biased andsever al groups of mo dern birds have in all probability ex isted in con­tinental areas (see belowj.r the recent studies have demonstrated a greatshar e of archaic (or ab errant) for ms (Wyleyia, Gobipteryx, Enantiornithes,Ichthyornithes, Hesperornithes) in the Cretaceous avifauna.

ECOLOGICAL TYPES

At least three ecological types can be broadly defined among thedescribed Cretaceous birds:

1. Piscivorous birds represented by seven (two undescribed) ge ner a ofHesp erornithes (see below) and two genera of Ichthyornithes (see below).Also two Ma estrichtian genera may fall into this category whether or notthey were correctly placed among the Pelecaniformes: Elopteryx fromthe Hateg (Hatszeg) Basin , Romania, and Graculavus, with t wo speciesfrom the Hornerstown Formation, New Jersey.

2. Shore birds represented by five genera of charadriiform birds, in­cluding Lonchodytes (Olson and Feduccia 1980a). Telmatornis m ay alsobelong to this category (Cracraft 1972). All they come from the Maestricht­ian formations of North America: Lance (Cimolopteryx, Palintropus,Ceramornis, Lonchodytes); Hell Cr eek (Cimolopteryx), Navesink ( Palaeo­tringa, Telmatornis) and Hornestown (Telmatornis).

3. Terrestrial birds represented by Gobipteryx from the Barun GoyotF ormation, Mongolia (Elzanowski 1977), and possibly also the Enantiorn­ithes, described by Walker (1981) from Lecho Formation, Argentina.Gobipte ryx has a robust, toothless beak, with strong edges, suggestive

. \

BIRDS IN C R E T A CE O US ECOSYSTEMS 79

of feeding on rather hard and easily available objects. Several embryonicskeletons, found in beds equivalent to Barun Goyot Formation, probablyalso belong to this genus (Elzano wski 1981).

Birds representing at least three other ecological types have beenclaimed to be found in the Cretaceous record. The flamingos, represent­ing filter-feeding waders, became renowned for the antiquity of theiralleged relatives, but this is probably a myth created by palaeornitho­logists (Feduccia 1980); the earliest unequivocal flamingo, based on suffi­cient material, is Juncitarsus fr om the middle Eocene (Olson and Feduccia1980a) and the oldest known filter-feeding wader is Presbyornis pervetusfrom the early Eocene (Olson and Feduccia 1980b). However, filter-feed­ing waders might ha ve arisen at ear lier date and the Eocene Presbyornismay be a temporal r elic (Feduccia 1977).

Harrison and Walker (1975) assigned to the owls (Strigiformes) twolarge tibiotarsi from the Hateg Basin, Romania. However, the strigiformresemblances are inferred rather than visible and the bones do notwarrant to be avian either (Brodkorb 1978).

Brodkorb (1976) described Alexorni s antecedens from the (!)marineBocana Roja Formation (Mexico) as a hypothetical ancestor of Piciformesand Coraciiformes. Seven fragments of postcranial bones display twentycharacters, which are unique to A lex ornis, and twenty one characterswhich are shared (or agreein g) with either Momotidae or Bucconidae, orboth. The most striking among the unique characters is the transverseposition of the external condyle of humerus, which approaches in thisrespect the condition found in Enantiornithes (Walker 1981). In functionalterms, a transverse ori entation of the external condyle suggests that theautomatic flexion of the manus - (as coupled with the humeroulnar fle­xion) was weak or nonexisting, which would be unusual among arborealbirds. These may have come into existence during late Cretaceous butwe have no firm evidence for their occurrence prior to the Eocene.

Conspicuously lacking in the Cretaceous record are diurnal raptors,herons (fish-eating waders) and ducks which are all mainly and primarilycontinental birds. Therefore some uncertainty remains about the possi­bility of their Cretaceous existence. Both ducks and herons are apparentlyvery likely to be preserved but the Cretaceous record of continental birdsmay be extremely biased. The oldest anseriform birds (Romainvillinaeand other small forms) and herons are mentioned from the late Paleoceneof central Asia (Kurochkin 1976).

SEA BIRDS

Almost all marine birds known from the Cretaceous are the toothedbirds, Ichthyornithes and Hesperornithes. By the end of Cretaceous themarine habitat may have been also exploited by the modern flying birds,

80 ANDRZEJ ELZANOWSKI

as suggested by the Maestrichtian occurrence of Elopteryx and Gracu­lavus in coastal plain deposits; and the beaches were certainly frequentedby shorebirds.

The toothed birds come predominantly almost exclusively from theregion of Western Interior seaway (or basin), which supported a richassemblage of fishes (mostly predatory) and other fish-eating vertebratesincluding the abundant birds. The vertebrate .fauna contrasts with theinvertebrate assemblages which are clearly impoverished as compared tothose of fully marine deposits , and several large groups are lackingaltogether; this is interpreted as evidence for special environmental con­ditions (Miller 1968; Kauffman 1977), not yet fully understood (?lowsalinity, ?oxygen shortage). On the other hand, the Hesperornithes at leastare not likely to have been common in other seas since their solid bonesand teeth, like those of cosmopolitan mosasaurs, would be certainly noticed.It appears therefore not unlikely that the toothed birds were attracted byopportunities of the epicontinental sea from inland waters where they hadexisted long before. After all, birds adapted probably first ~o inlandwaters and only thereafter some of them could invade seas.

Both the Ichthyornithes and Hesperornithes certainly fed largely onfishes; one of the coprolites referrable to Baptornis, contained a jaw ofthe salmoniform fish Enchodus (Martin and Tate 1976). An additionalsource of food may have been provided by cephalopods, especially thesquids which are not rare in the Smoky Hill Member of the NiobraraFormation (Miller 1968); squids constitute an important source of foodfor the modern sea bird communities (Ashmole and Ashmole 1967).

The Ichthyornithes are known only from the marine deposits of thesouthern half of North America (fig. 1), mostly from the Niobrara For­mation, Kansas. Ichthyornis is being commonly compared with a tern,what may suggest catching of prey by surface plunging or dipping (Ash­mole 1971). This comparison, however tempting as it is, may be mislead­ing (e.g. in restoration of feeding habits) since the osteology of flightapparatus of Ichthyornis is very different from that of any modern birdand the differences are not understood functionally; if the flight was asdifferent as the morphology is, this comparison would be hardly tenable.

Apatorn is celer, known from a few remains from the Niobrara Forma­tion, is usually placed in the Ichthyornithes, but it needs restudy (Brod­korb 1971). Howard (1955) noticed that "the available elements of Apa­tornis are so different from those of its famous contemporary [Ichthyornis]that it is entirely possible that the two are members of distinct or­ders".

The He sperornithes include six late Cretaceous genera (Hesperornis,Coni ornis , Baptorni s, Neogaeornis and two undescribed) and probablyalso the Albian genus Enaliornis (Martin and Tate 1976, Martin 1980).

Except for Neogaeornis known from a single bone (tarsometatarsus)

BIRDS IN CRETACEOUS ECOSYSTE MS 81

Fig. 1. Rec or ds of Hesperornis (1- 7, circles) and Ichthyornis (7- 10, triangles). (1 )Ignek Formation, Alas ka : one spe cime n, species inde te rminate (M. Brett-Surmanand La ur ie McDonald, pers. cornm.). (2) "Brow n Beds" Northwest Territories :remains of 10 indiv iduals of H. regati s (Ru ssell 1967). (3) Foremost Formation, Al­berta : one specimen of H. cf. regatis (Fox 1974). (4) Vermilion Formation, Boyneand Pembina members, Ma ni toba: four specimens of H. rea al i s (Bardack 1968). (5)Claggett Formation (th e upper part) or Judith River Formation (the basal p art) ,Montana: two specimens of Conior nis altus w hic h is recen tly considered to be con­generic with Hesperornis (Fox 1974, Ma rtin 1980). (6) Pierre Shale (th e lowe r par t),So uth Dak ota : possibly as many as fo ur spe cies, one of the m inclu di ng a ll larger ,r ega tis-sized f~ms (Martin 1980). (1) Niobrara Formation, Smo ky Hill Member ,Kansa s: Hes pe ror71is: in stead of th e th ree species descri bed by Marsh there areprobably on ly two, H. regatis inc luding all la rge forms and an undescr ibed speciesincludin g sma ller for ms (Martin 1980); by the time of La mbrecht (1933) the Yalecoll ections inc lude d th e remains of about 50 indivi dua ls refe rr ed to H. regalis andH. crass ipes. Ichthyorni s: the min im um n umbe r of inviduals, as inferred fro mMarsh's (1980) monogr aph, is fifty seven ; they are refer r ed to as many as six (I)species (Brodkorb 1967). (8) Carlile Formation (base) or Greenhorn Formation (top) ,Kansas : one specimen, species ind eter minate (Walker 1967) . (9) Austin Chalk, Texas:one specimen referred to I. !entus (Brodkorb 1967). (10) Selma Chalk, Alabama: one

specimen of I. an tecessor (Olson 1975).

6 Acta P a laeontologlca Polonica N r 1-2/83

82 ANDRZEJ ELZANOWSKI

from the Quiriquina beds, Concepcion Province, Chile, all the Late Cre­taceous Hesperornithes come from the western half of North Americabetween Kansas and the latitude 69°N (fig. 1). The most common is thegenus Hesperornis, reported to be especially abundant in .the "BrownBeds" (Russell 1967) and the Vermilion River Formation (Bardack 1968).In the Niobrara Formation the relative abundance is lower but a highabsolute number of specimens testifies that the birds were common in thisarea. The youngest Hesperornis fossils are those from the Campaniandeposits of the Foremost Formation, Pierre Shale and Pembina Memberof the Vermilion (Fox 1'974). "Brown Beds" and the Boyne Member of 'th e 'Vermilion are thought to be approximaly isochronous to the Smoky HillMember, which, according to Kauffman (1977), extends from the lateConiacian to the earliest Campanian.

Other hesperornithiforms from Niobrara Formation are Baptornis ad­venus known from more than ten individuals (Martin and Tate 1976)and two undescribed genera (Martin 1980). New nonmarine hesperornithi­forms have been recently announced from the Maestrichtian deposits ofSouth Dakota (Martin 1980). '

The Hesperornithes covered a size range between larger grebes andlargest living penguins. Like the penguins, they are good indicators ofa high productivity of marine waters. The diving achievements of largerhesperornithiforms, as Hesperornis . regalis or Baptornis advenus, musthave been significantly greater than those of much smaller loons anddiving ducks which are known to reach 55 m of depth (Kooyman 1975).Thus the larger hesperornithiforms could probably reach a bottom dwell­ing prey - at least in shallow areas like the Niobrara sea, which waspresumably about 40 m deep during Smoky Hill Member deposition (Miller1968).

In spite of being flightless , the Hesperornithes apparently had to co­exist with the isurid (Isurus, Lamna, Squalicorax) and carchariid (Sca­panorhynchus) sharks, as well as many large teleost predators (Lane1944; Bardack 1968) including the huge Xiphactinus audax. It is thereforequite probable that for breeding and raising their young they had to seekprotected sites in coastal areas (Fox 1974).

The abundance of great flightless and foot-propelled divers in thewarm seas of the Cretaceous constitutes a striking difference in com­parison with the Cenozoic seas. Tertiary seas supported only wing pro­pelled divers which, with a minor exception of the Pliocene mancallas(Howard 1970) extending to the Baja California (Mexico), fl ourished mai­nly in the temperate seas; this applies first of all to the largest Tertiarypenguins and plotopterids which vanished in the Miocene, being pro­bably displaced by porpoises and seals (Stonehouse 1969; Simpson 1975 ;Olson and Hasegawa 1979; Olson 1980). In the Quaternary seas theflightless wing propelled divers (penguins and auks) appear to have been

BIRDS IN C RET A CEO US E C O S YS TEMS 83

ousted toward the polar zones . The modern communities of sea birds intropical and subtropical zones ,consist almost exclusively of terns, gulls,pelecaniforms and procellariiforms (Bourne 1963; Ashmole 1971), whichare all good flyers. Most of them ' (especially the smaller ones) dependpartly or entirely on schools of pr~datory fish, mainly tunas, to drivetheir prey closely to the surface and so make it available (Ashmole andAshmole 1967). In tropical water s , the av ian pursuit divers would pro­bably find the ' food too deep or too sca ttered for enconomical hunting(Br own 1976) and, at the same time, would have to compete with largeand swift perciform pr edator s as the scombroids ' (including the tunaswhich are partly homeotherms!), the sphyraenids (including the bar­racudas) and several percoids. Thus, the r ise of large perciform predators ,must have been of crucial impor tance for the evolution of sea bird 'com­munities.

It would ,be surprising if the profound change in the fish faunas to­wards the close of Cretaceous had no influence on the communities ofpiscivores. The acanthopterygians, which fi rst appear in the Cenomanian,remained almost constant in numbers until the end of the period (Patter-

\

son 1964) and the fish fa unas were strongly dominated by the lower tele-osteans ; in the Niobr ara Formation the only acanthopterygian fish isthe-small (7 em long) Kansius sternbergi , a holocentrid berycoid (Patter­son 1964). By the end of Cretaceou s the lower teleosts su ccumbed to thepressure of competition from the beryciforms and their derivatives (Goody1969) especially the perciforms which bega n to radiate into a great .rangeof forms (Patterson 1964); large and swift predators have apparentlyarisen very early since the first tunas (Palaeothunnus) are known fromthe Paleocene (Bannik ov 1978). Being well protected by specialized spines,the acanthopterygians, and es pecia lly the perciforms (Marshall 1971) dwetheir success to th e r adical improvement in locomotion , especially themanoeu vrability, and feeding techniques, and therefore their impacton the fish-eating bir ds and reptiles must have been twofold: the pisci­vores became subject to powerful competition and, at the same t ime,deprived of the us ual prey they were adapted to fee d on, since the low erteleosts (e.a. myctophoids) had been displaced to deeper sea layers (Mar ­shall 1971). The explosive r adiation of aca nthopte rygians appe ars thusto be an important fa ctor of exti nction of both the marine reptiles andtoothed birds alt ho ugh the process may have involved also other factorsand the dis appearance of the Western In terior basin may have stro nglyreduced the pop ulations of toothe d birds.

Oelofsen (1978) ex te nded on Hesperornis his elegant h ypothesis of0 2/COz imbalance as the cause of ex tinction of lar ger, homeothermic andoviparous vertebrates. Yet not all Hesp erornithes we re la rge and alsothe Ichthyornithes disappeared although they were small . On the otherhand , sma ll birds may be also affected by Oz/CO z imba lance if the y

G'

84 ANDRZ EJ ELZANOWSKI

bury their eggs as do the recent megapodes. As usual, there are too muchof possible extinction factors and the competing hypotheses are hardlyverifiable.

TERRESTRIAL BIRDS

Aside from Gobipteryx and Enantiornithes, th ere are good reasons tobelieve that the terrestrial birds had been significantly diversified dur­ing the Cretaceous (Brodkorb 1971 ; Mayr 1976). The prime candidatesfor the Cretaceous gr ou nd-feeding birds are small flying ancestors ofthe modern palaeognathous birds which include the tinamous and ratites,and appear now what they were thought to be by Thomas H. Huxley i.e.the "waifs and strays" of an old (Cr etaceous) radiation (Prager et al.1976 ; Simpson 1980). Among m odern birds the palaeognaths retained themost primitive characters in the structure of skull and pelvis as well asin the reproductive biolo gy (Elzanowski in prep.). These characters,shared by small and flying tinamous, have nothing to do with eitherflightlessness or giantism, and the claim for their neotenic origin isa pure ad hoc h ypothesis . The gru ifor m origin of the ostrich, recentlypostulated by Feduccia (1980) is unacceptable sinc e the ostrich's palae­ognathous ch aracters , which cons picuous ly do not appear in any neogna­thous fli ghtless bird, would be then unexplained. The present distributionof thepalaeognaths al so stro ng ly suggests the Mesozoic occur r en ce of theira ncestors which may be sa fely assumed to have been flying n ot betterthan do the tinamous and ther efor e u nable to cross se a barriers (tina­m ous are apparent ly unable to colonize any neotr opical isl and). Thus theh ypothesis that the ratites are mon ophyl etic and derived from a fl ightlessancestor, as sta te d r ecently by Cracraft (l974) and challenged by Feduc­cia (1980) is unnecessary to postulate the Mes ozoic disp ersal. P alaeogna­thous birds ar e r emnants of a prim it ive grade of avia n ev olution charac­terized i.a . by a limit ed power of fl igh t. Most simila r to the ancestralpalaeognaths are certai nly the Sou th Ame ri can t inamous, wh ich are shortdistance fl yers, wi th a cons pi cuo us ly defficient manoeuvrability. Tina­mous are otherwise ga ll ifor m in habits, appearance and str ucture of thepostcranial ske le ton (except for the pel vis).

Prim it ive ga ll iforms are al so very likel y to h ave occurred in the Cr e­taceous. Cracr aft (1973a, 1980) argu es for the Gondwanian dispersal ofthe m egapodiids and cr acids, but Rich (1975) concluded that the Indoma­la yan and Antarcti c r outes are eq ually likely for the m egapodes. Nowthe prirrlitive position of the magapodes appears to be better substan­tiated (Prager and Wilson 1976), and espe cial ly their biology of repro­duction, in spite of some sp ecialized fea tures , may have ancient origins(Elzanow ski in prep.). If the megapodes are considerably older than the

BIRDS IN CRETACEOUS ECOSYSTEMS 85

phasiands ' which appeared in the Eocene, the Antarctic route remainsthe only plausible possibility, since Australia approached Asia only duringthe Eocene. Bearing in mind the galliform similarities of the tinamous, itis perhaps worth noting that Prager and Wilson (1976) suggested the gal­liform "phenotype" as one of the most primitive types among the neogna­thous terrestrial birds.

Birds of unquestionable gruiform-ralliform affinities are not knownbefore the early Eocene, but they appear then in such a great diversitythat Cracraft (1973b) was probably right in extending their ancestry tothe Cretaceous. In any case the small cursorial ancestors of diatrymas,known from the Paleocene and commonly thought to be of gruiformaffinities, have been in all probability present in the Cretaceous.

Ostrom (1974, 1976) has convincigly shown that there is no evidencefor the arboreal habits of Archaeopteryx and all other premises suggest,though indirectly, that the terrestrial habits are primitive for birds:(1) Almost all avian taxa that may be reasonably supposed to be pri­mitive include either terrestrial or water birds. Cracids are a major ex­ception but they do not appear to be highly specialized in the arboreallife either in morphology or in nesting habits; they have strong legsand feed partly on the ground. (2) The Cretaceous embryos (Elzanowski1981) confirmed a generally accepted view that the precocial/nidifugousmode of development is primitive for birds. Moreover, most primitiveamong modern birds are predominantly ground nesters (the cracidsexcepted) and have precocial chicks; both of these premises strongly sug­gest that the arboreal nesting appeared rather late in the avian evolu­tion. (3) Paradoxically, the Albian record of the foot-propelled divers(Enaliornis) constitutes the best evidence for an early radiation of ter­restrial birds since the foot-propelled divers are likely to originate onlyfrom birds that had once walked on land, then began to do it in water,and finally adapted to swimming.

SCAVENGING REPTILES AND TAPHONOMIC IMPLICATIONS

The strong dominance of water-dependent birds in the Cretaceousrecord certainly represents a bias which ~esults from (1) the higher fossi­lisation potential of aquatic organisms, a poirit that requires no furthercomments, (2) the poor knowledge of truly continental assemblages, andpos sibly also (3) a specific taphonomic factor acting against the preser­vation of terrestrial birds.

In those early Cretaceous formations that are well studied, the re­mains of land vertebrates are scanty and mostly fragmentary (Ostrom1970) with the major exception of Bernissart in Belgium, whereas themost promising vertebrate assemblages, especially those of central Asia,are as yet poorly known.

86 ANDRZEJ ELZANOWSKI

Late Cretaceous assemblages of land vertebrates are mostly recoveredfrom coastal plain deposits, where the dominance of the water-dependentbirds is to be expected. The major exception are the Mongolian forma­tions, and one of them (Barun Guyot) yielded' Gobipteryx, in ,all pro­bability a true terrestrial bird. If, however, terrestrial birds were asdiversified as implied above, the puzzling question is, why only one birdspecies was found in Mongolia, while other small vertebrates as, lizardsand mammals are represented by numerous taxa (see Osm6lska 1980)?

A remarkable feature of the Cretaceous record is the rarity of mediumsized vertebrates, especially those of less than 10 kg in weight. Thisapplies both to the medium sized species and the juvenile specimens oflarger animals, especially dinosaurs. The lack of small dinosaurs ledBakker (1971, 1972) to the known conclusion that the small dinosaurscould not exist because of the thermic requirements. Aside from thecontroversy about the homeothermy of dinosaurs, this explanation doesnot account for the known scarcity of juvenile dinosaurs, which is be­lieved by Richmond (1965) to result from the high mortality of juvenilesand the dominance of adults in the dinosaurian populations. This howeverraises the question of why there is no abundant fossil evidence of sucha high juvenile mortality (unless it is assumed that death occurred mostlyat earlier embryonic stages where the skeletons are little ossified). Smallervertebrates may be underrepresented in the skeletal remains as shownby Behrensmeyer et al. (1979) for the African savanna. However, themicrovertebrates (lizards, mammals) are well represented in the Creta­ceous record whereas the smaller among medium-sized specimens areunexpectedly rare.

A major factor, possibly responsible for the lack of medium sizedvertebrates in the fossil record, is scavenging and predation by carni­vorous reptiles, mainly theropod dinosaurs. The feeding method of largereptilian carnivores is essentially different from that of mammalian pre­dators. Both crocodiles and large lizards (monitors) swallow the itemsas large as can pass through their gullets. Auffenberg (1972) noticed thatthe Komodo dragons (Varanus komodoensis) swallow bones even whencleaned. of flesh (but not too dry) and the skeletal remains of dragonsare said to be rare on the island because of scavenging by their conspe­cifics (Peters 1963).

To the contrary, the carnivorous mammals, including the largest (asthe lion) tend to disarticulate the larger prey and eat only the selectedparts, discarding the others, as heads or distal parts of the legs (or wings).That is why skeletal remains are very common in the wild are.as of opencountry as the central Asiatic steppes or the African savanna. Moreover,the mammalian carnivores commonly cache the remaining parts of prey,mainly in the ground, and bring the prey to their young. Both habitsmay largely contribute to the dispersion of bones.

BIRDS IN CRETACEOUS ECOSYSTEMS 87

The large reptilian carnivores, both the crocodiles and monitors, arevery fond of carrion in any stage of decay. Similar feeding habits maybe safely assumed for the carnivorous theropods. In contrast to predationwhich is size selective, scavenging is largely non-selective with respect tosize , i.e. scavengers take any carrion they perceive. Therefore, assumingthe same detectability by scavengers, smaller carrion has a greater chancethan the larger one to be swallowed whole, since both smaller and greaterscavenger can do it. In fact, the detectability of carrion may stronglydepend on its size and the cadavers of microvertebrates, lying in vegeta­tion or various natural crevices, are certainly less detectable than thoseof larger animals especially when sight-directed scavengers are active.The greater chance for the smaller carrion to be swallowed whole andthe smaller chance to be detected are two conflicting factors acting join­tly against preservation of medium sized animals. I would predict thatthere is a critical medium size range of animals, small enough to be ·swallowed whole, and large enough to be well detectable and/or acces­sible. The detectability of carrion will be also influenced by the behaviourof living animals e.g. byburrowing habits which will increase the chanceof preservation of skeletons among fossorial forms. I

A few records of specimens approaching the critical medium size mayconfirm the presented hypothesis after careful palaeoecological considera­tion. For instance, juvenile dinosaurs (protoceratopsians), represented bynearly all growth stages, are known to be abundant only in Bayn Dzak,Djadokhta Formation, Mongolia and the same locality yielded also thesmallest specimen of a theropod (ZPAL MgD-II/29, under study by theauthor) with the skull length of about 5 em. In accordance with the pro­posed hypothesis, larger carnivores are scanty in the Djadokhta Forma­tion, especially in Bayn Dzak (Osm6lska 1980). Aside from dinosaurs,other Cretaceous medium sized land vertebrates are large lizards fromthe families Varanidae, Polyglyphanodontidae and Macrocephalosauri-

. dae, which, however, may have been partially burrow dwellers and/orsemiaquatic in habits (Estes 1964) as are the modern varanids (Auffen­berg 1980).

Bird cadavers appear to be more conspicuous than those of most othervertebrates of comparable size because of feathers which are oftenspread out, moved by the wind and, at later stage of decomposition, dis­persed around the cadaver. These may be important cues in the search­ing image of the scavenger. Even when falling in water, the bird cadaversare not immune from scavenging by land or flying vertebrates since,contrary to other vertebrates, they float for a long time at the surface,'as shown by Schafer (1962).

Of prime importance for my hypothesis is the ability of reptiles todigest bone. Snakes may digest bone completely, although small frag­ments of bone sometimes occur in the excrements (Skoczylas 1978). Bo-

88 ANDRZEJ ELZANOWSKI

nes were also found in the fecal pellets of Varanus komodoensis byAuffenberg (1980). Crocodiles are known to decalcify bones and teethcompletely although the organic matrices may be left intact (Fisher 1981and references therein). In contrast, the mammalian digestion of bone islargely incomplete because food passes through the alimentary tractmuch more rapidly. This is due to the hi gh levels of body temperatureand metabolism in mammals. Dinosaurs may be considered as home­otherms, but this does not imply a mammalian level of body temperatureand metabolic rate (Skoczylas 1980; Spotila 1980) and therefore the rateof food passage in dinosaurs may have been similar to that of otherreptiles. Swinton (1970) remarked upon the absence of any solid materialsin dinosaur coprolites and suggested that the digestion of bones was aseffective as that of crocodiles.

If the presented explanation of the rarity of medium sized vertebratesin the late Mesozoic is at least partly true, both the mode of feeding andthe digestive rates of reptilian predators, which differ from those ofmammals, must impose a significant limitation on the actualistic approachto the taphonomy of Mesozoic vertebrate assemblages.

Acknowledgements. - I thank P . Brodkorb (University of Florida), R. Molnar(Queens land Museum, Australia) an d C. A. Walker (British Mu seum) for the com m ­ents and unpublish ed information. M. Brett-Surman (George Washington Univers­ity), Laurie McDonald (University of California) and D. McInnes (Provincial Museumof Alberta) gen erousl y provided locality data for occurr ences of Hesperornis.J . S. Mellett (New York University) , A. Seilacher (Ge olo gisches Institut, Tiibingen).and S. Tarsitano (Qu eens College, Flushing) kindly improved the early drafts.W. Cl emens (University of California ) and A. J erzrnanska (Uniwersytet Wroclawski)made va luab le suggestions concern ing taphonomic and pala eoichthyolog'ical ques­tions, r espectivel y. P . Currie (Provincial Mu seum of Alberta) placed at my disposalthe m anuscript on the fo otprints of Cretaceous birds. P. Currie, P . Dodson (Universityof P ennsylvania) , R. Ma r cinowski (Uniwersytet Wars zawsk i) and H. W. Miller (FortHays Kansas State College) pr ovided m iss ing literatur e.

REFERENCES

ASHMOLE, N. P. 1971. Sea bi rd eco logy and the marine environment. In: D. S.Far­n er an d J . R. King (eds.), Avian Biol ogy 1, 233-286. A cademic Press, NewYo rk and L ondon.and ASHMOLE, M. J . 1967. Com parative feeding ecology of se a birds ofa t r opi cal oce anic isl and. - B ull. Peabody Mus. Nat. Hist., 24, 1-131.

, AUFFENBERG, W. 1972. Komodo dra gons. - Natur. H istory, 81, 52- 56.- 1980. The herpet ofauna of K om od o, w it h n ot es on adjacent areas. - Bull.

Flori da St ate M us ., Biol. Sci ., 25, 2, 39-156.BAKKER, R. T. 1971. Dinosaur physiology and the origin of mammals. - Ev olu t i on ,

25, 4, 636-658.

BIRDS IN CRETACEOUS E COSYSTEMS 89

- 1972. Ana tomica l and ecologica l evide nce of en dothermy in dinosaurs. ­Nature, 238, 81-85.

[BANNIKOV, A. F .] 1978. BAHHI1KOB, A. <P. HOBbH'1: pozt cKyM6pHeBblx H3 aepxnerorraneouena TYPKMeHHH. - Tla.neowr . lKypn. (1978), 4, 79-83.

BARDACK, D. 1968. Fossil ver tebra tes f r om the ma rine Cretaceous of Manitoba.­Can. J. Eearth sa; 5, 145-153.

BEHRENSMEYER, A. K., WESTERN, D. an d DECHANT BOAZ, D. E. 1979. Newperspectiv e in vertebrate pa leoecology fro m a recen t bon e assem b lage.­Paleob iology ,S, 1, 12-21.

BONAPARTE, J . F . and POWELL, J . E. 1980. A continen ta l assem blage of te t rapodsfrom the Upper Creta ceous beds of EI Br ete, nort h- wester n Argentina (Saur­opo da - Coelurosauria - Carnosauria - Aves). - Me m . Soc. geol. France, N.S., 139, 19-28.

BOURNE, W. R. P . 1963. A revi ew of oceanic stud ies of the bi ology of seabirds.­P r oc. 13th Int. Om. Congr., 831-854.

BRODKORB, P. 1967. Ca talogu e of fossil birds: pa rt 3 (Ra ll iformes, Ichthyornithi­formes, Charadriiformes). - Bul l. Flor i da St at e Mus., BioI. Sci., 11, 3, 99-220.1971. Origin and evolution of birds. I n : D. S. Farner and J . R. King (ed s.),Avian Biol ogy, 1, 19-55. Acade m ic Press, New York and London.1976. Discovery of a Cr etaceous bird, apparen tly ancestral to the orders.Coraciiformes and Picifor mes (Aves : Carinatae). - Sm i thson ian Contr. Paleo­biol ., 27, 67- 73.1978. Ca talogu e of foss il b irds, par t 5 (Passeriformes). - Bul l. Florida Stat eM us. , B i oI. Sci. , 23, 3, 139-228.

BROWN, R. G. B. 1976. S eabirds of S ou th America a nd the North West Atlan tic . ­Proc. 16th Int. Orn. Congr., 716-724.

CASAMIQUELA, R. M. 1980. Considerations ecologiques et zoo geograph iques sur lesVe r tebres de la zone littorale de la mer du Maastrichtien da ns le Nord de laPatagonie . - M em. Soc. geol. France, N.S ., 139, 53-55.

CRACRAFT, J. 1972. A new Cretaceous charadriiform fa mily. - Auk, 89, 36--46.1973a. Continental drift , paleoclimatology, and the evolution and bi ogeo­graphy of bi rds. - J. Zool. Lond., 169, 455- 545.1973b. Systematics an d evolu tion of th e G ruiformes (Cl ass Aves), 3: P hyl ogenyof the suborder Grues. - BuH. A m. Mus. N aiu r . R i st. , 151, 1, 1-127.1974. Phyl ogen y of evolu ti on of ra ti te b irds. - Ib i s, 116, 494-521.1980. Av ian phyl ogeny a n d intercon ti nenta l biog eogr aphic pa t terns. - Proc.17th Int. Orn. Congr., 1302- 1308.

CURRIE, P . J . 1982. Bird footprin ts from the Gethi ng F ormation (Apti an, Lo werCretaceous) of northeastern British Columbia , Canada . - J. Vert . Palaeont.,1, 3- 4, 257-264.

ELZANOWSKI, A. 1977. Skulls of Gob ip t eryx (Av es) from the Upper Cretaceo us.of Mongolia : I n : Z. Kielan-Jaworowska (ed .), Results Pol.-Mong. P al aeont.Exped s., VII . - Palaeont . Polonica, 37, 153-165.1981. E mbryon ic bird skele tons f ro m the .La te Cr etaceous of Mongoli a . I n :i bidem, IX. - I bi dem , 42, 147-179.

ESTES, R. 1964. Fossil vertebrates from the L a te Cretac eous La nc e F orma tion.Eastern Wyo m in g. - Uni v. Calif . Publ. Geol. Sci. , 49, 1-180.

FEDUCCIA , A. 1977. H ypothetical stages in the evolu tion of m od ern ducks a n dfl amingos. - J. T heor. BioI., 67, 715-721.

- 1980. The Ag e of Bird s. 196 pp. Harvard Un iv. P ress, Cambridge (Ma ss.) andL ondon.

FISHER, D. C. 1981. Cr ocod ili an sc a to logy, m i.cr ov er tebra ta con centra ti on s a n dename l- less teeth . - Paleobi ol ogy , 7, 262- 275.

90 A N D R Z E J E LZAN O W S K I

FOX , R. C. 1974. A Middl e Ca m panian , nonmarine occurrence of the Cretace ousto othed bird H esperorni s Ma rsh . - Can. J. Earth Sci., 11, 9, 1335-1338.

GOODY, P. C. 1969. The r elationships of certain Upper Cretaceous teleosts withspecia l r eference to the myct ophoid s. - Bu ll. B r i t. Mus. Nat. H i st. (G eol.),Suppl, 7, 1-255.

GRADZINSKI, R. and JERZYKIEWICZ, T. 1974. Sedimentation of the Barun G oy otFormation. In: Z. Ki el an-Jaworow sk a (ed .), Results Pol-Mong, Palaeont. Ex­peds., V. - Palaeont. Polonica, 30, 111-146.

GRIGORESCU, D. and KESSLER, E. 1980. A new spec imen of Elopte ryx nopcsaiAndrew s from the Dinosaur ia n Bed s of Hateg Basin . - Rev. Roum. Geot.Geophys. Geogr. , Geol., 24, 171-175.

HARRISON, C. J . and WALKER, C. A. 1973. W yley ia: a new bird humerus from theLower Cretaceous of England. - Palaeontology, 16, 721-728.

- and - 1975. The Bradycn ernidae , a new fa mily of ow ls from the Upper Creta­ceous of Romania. - Ibidem, 18, 563-570.

HOWARD, H. 1955. A new wadi ng b ird from the Eoc ene of Patagonia. - Amer.Mus. Novit. 1710. 1-25.

- 1970. A r eview of the extinct avian 'gen us, Mancalla. - Contr. Sci . Nat . Hist.Mus. Los Angeles County, 203, 1-12..

JENSEN, J . A. 1969. Fossil eggs from Utah and a concept of surviving featheredr eptiles. - Proc.• U tah A cad . Sci . Art s L etters, 46, 1, 125-133.

KAUFFMAN, E. G. 1977. Geological and b iological overview: Western InteriorBasin. - Mountai n Geologist, 14, 75-99.

KOOYMAN, G. L. 1975. Behaviour a nd physiol ogy of diving. I n : B. Stonehouse(ed .), The Biology of P engu in s, 115- 137. Ma cmillan Press, London and Ba­singstoke.

KUROCHKIN, E. N. 1976. A su rvey of the Paleogene birds of Asia. - SmithsonianContr. Paleobiol., 27, 75-86.

LAMBRECHT, K. 1933. Handbuch der Palaeornithologie. 1024 pp. G ebr. Borntraeger,Berlin.

LANE, H. H. 1944. A survey of the fo ssil vertebrates of Kansas. Part 1: The fishes. ­T rans. K an sas A c. Sci. , 47, 129-176.

MARSH, C. C. 1880. Odontornithes: Mon ographs on the Extinct Toothed Birds ofNo r th America. - US Geo!. Ex pl oration of the Fortieth Parallel, 7, XV+201pp., Wash ington . .

MARSHALL, N. B. 1971. Explorations in the Life of Fish es. Harva rd Univ. Press,Cambridge (Mass.).

MARTIN, L. D. 1980. F oot-pr opelled divin g birds of the Me sozoic. P roc. 17th Int.Orn. Congr., 1237-1242.

- a nd TATE, J . 1976. The skel e ton of Ba ptornis advenus (Aves: Hesperornithi­fo rmes). - Smithson ian Con t r . Paleobiol ., 27, 35-66.

MA YR , E. 1976. [A contribution to the d iscussion on the or igins of Austral as ian avi­fauna] . - Proc. 16th Int. Orn. Congr., p. 98.

ME HL, M. G. 1931. Addition s to the verte bra te record of the Dakota Sandstone. ­Am. J . Sci ., se r. 5, 21, 441-452.

MILLER, H. W. 1968. Inve r teb rate fauna an d en vironment of deposition of the Nio­brara Formation (Cretaceous ) of Kansas. - Fort Hay s Studi es, N. S., Sci. S.,8, 1-'-90.

M OLNA R, R. 1980. Australian La te Mesozoic terrestria l tetrapods: some in ter pr et a ­ti ons. - Mem. Soc. geol . France, N. S. , 139, 131-143.

OELOFSEN, B. W. 1978. Atmos pher ic carbo n dioxide/o xygen imba lanc e in the LateCretac eous , hatching of eggs and the ex ti nc tion of bi ota. - Palaeont. africana,21, 45-51.

BIRDS I N C RETACEOUS E COSYSTEMS 91

OLSON , S. L. 1974. [A rev iew of] Sys tematics a n d evolu tion of t he Gruiformes, byJ oel Cracra ft . - The A uk, 91, 4, 862-865.1975. Ichthyornis in the Cretaceous of Al abama. - Wilson BulL, 87, 103-105.1980. A p ew genus of pengu in-like pe lecaniform bi rd from the Oligo cene ofWashington (P el ecaniform es: Plotopteridae). - Cont r . Sci . Natur . R i st . Mus.Los A ngele s COUT!ty , 330, 51-57.a n d FEDUCCIA . 1980a. Relation sh ip s an d evolution of flaming os (Aves:Phoenicopteridae). - Sm ithson i an Contr. Zoo I ., 316, III +73 pp.and - 1980b. Presbyornis a nd the origin of the Anseriformes (Aves : Chara­d r iomorphae). - I bidem , 323, IV +24 pp.a nd HASEGAWA, Y. 1979. Fossil counterparts of giant penguins from theN orth P acific. - Sci en ce, 206, 4419, 688-689.

OSMOLSKA, H. 1980. The Late Cretaceou s ver tebrate assemblages of the GobiDesert , Mongoli a . - Mem. Soc. geol. France, N. S. , 139, 145-150.

OS TROM , J. H. 1970. S tratigraphy and paleontology of the Cl ov e rly Formation(Lo wer Cretaceous) of the Bighorn Basin Area, Wyoming and Montana. ­Bull. Peabody Mus. N at . R i st. , 35, 1-234.1974. Archaeopter yx a n d the origin of flig h t. - Quart. Rev. BioI., 49, 27­47.1976. Arch aeop ter yx a nd th e origin of bi rd s. - B ioI. J. L in n. Soc., 8, 91­182.

PATTERSON, C. 1964. A rev iew of Mesozoi c acanthopterygian fi sh es with spe cia lr eference to those of the English Chalk. - Phil. Trans. Roy. Soc. London, B,247, 213-482.

PETERS, G. 1969. Klasse Reptilia - K ri echtie re. I n : Urania Ti erreich, F ischeLurch e - Kriechti ere, 355-507. Urania VerI., Leipzig- J ena-Berlin.

PRAGER , E. M. an d WILSON, A. C. 1976. Congruency .of ph ylogeni es derived fromdiffe rent proteins. - J . Mol. Evol., 9, 45-57., - , OSUGA, D. T. and FEENEY, R. E. 1976. Evolution of flightless landbirds on so u the rn continents : t ransferrin co m parison sh ows monophyleticor igin of ratites. - I bidem , 8, 283-294.

RICH, P . V. 1975. Antarctic d ispersal ro u tes , wandering continents, a n d the or iginof A us tralia's n on-passerifo rm avifauna. - Mem. Nat. Mus. V ictoria, 36, 63­126.1976. The hi story of bi rds on the island con ti ne n t Australia. - Proc. 16th Int.Orn. Congr., 53-65.and BOHASKA, D. J. 1976. The wo rld 's oldes t ow l: a new strigiform fromthe Paleocene of sout hwes tern Colorado. - Sm ithsonian Contr. PaleobioL,27,87-94.

RICHMOND, N. D. 1965. Perhaps juvenile dinosaurs w ere al w a ys sc a rce. - J . Pa ­l eont., 39, 503-505.

RUSSELL, D. A. 1967a . C retaceous v erteb rates fr om the Ande rson River, N.W.T. ­Can. J . Ear t h Sci., 4, 21-38.

SCHAFER, W. 1962. Aktuo-Palaontologie nach Studien in del' Nordsee 666 pp.W. Kramer, Frankfurt/M.

SCHLEE, D. 1973. Har zk onservierte fo ssile Vog elfed ern aus del' unte rsten Kreide. ­J . Orn. ; 114, 207-219.

SIMPSON, G. , G. 1975. Fossil penguins. In: B. Stonehouse (ed.) , The Biology ofPenguins, 19--41. Macmillan Press, London and Bas ingstoke.1980. Fossil birds a n d evolu ti on. - Contr. Sci. N at . R is t . M us. Los A ngel esCounty, 330, 3-8.

92 A N DRZ E J E LZAN O W S K I

SKOCZYLAS, R. 1978. Physiology of d igesti ve tract. I n: C. Gan s and K. A. Gans(ed s.), Bi ology of the Re ptilia , 8, Physiology B, 589-717. Academic Press, Lon­don and New York.1980. Strategie te rmoregu lacy jne ga d6w k opalnych [Thermoregula t ory st r a­te gies in the fos sil r ep tiles. In P oli sh, w ith English summary] . - Pr zeglqdZoologi czn y, 24, 1, 15-46.

SPOTILA, J . R. 1980. Con straints of body size and environ men t on the temperatureregu la tion of din osaurs. In : R. D. K. Thom as an d E. C. Ol son (ed s), A ColdLook a t the War m-Bl ood ed Dinosaurs, 233-252. AAAS Sel ect ed S ymposium ,28. Westvi ew Press, Bou lder, Color ado .

STONEHOUSE, B. 1969. Enviro nm ental t emperatures of Ter t iary penguins. - Scien­ce, 163, 673-675.

SWINTON, W. E. 1970. The Din osaurs. 331 pp, Allen and Unwin, L ondon .TALENT, J . A., DUNCAN, P . M. an d HANDBY, P . L. 1966. Early Creta ceous feathers

fr om Vict or ia . - Emu, 66, 81-86.WALKER, C. A. 1981. New subclass of bi rds fro m the Cr etaceou s of Sou th Ame­

rica . - N ature, 292, 51-53.WALKER, M. V. 1967. R eviv a l of in terest in the t oothed birds of Kansas. - T r ans .

K an sas Acad. Sci., 70, 60- 66.

NOTE ADDED IN PROOF

New recordLower Cre taceous of Asia. K uroch k in (1982: Dokl. Akad. Nauk SSSR, 262, 2, 452­

455) ga ve a prelim in a ry descript ion of an unquestion able carin ate, Ambior t us de­me ntjev i, from Neocom ian beds (sta ge and formation undet ermined) of Mongoli a(loca li ty: Khurilt-Ulan-Bulak). Mor eover he an no unced a second av ian skeleto n froma no t her locality of Mongolia a nd numero us prints of feat he rs fr om eigh t loca li t iesof Mongolia an d t ransb a ik a lian S iberia.

Upper Cretaceous of N or t h America. Lucas and Sullivan (1982: J. P a leont. 56,2,545-547) descr ihed a humerous of Ichthy or nis sp. from the late Turonian beds ofMancos Shale, Juana Lopez Mem ber, n or thw est ern New Mex ico; this r ecord exten dst he r ange of I chthyorn i s westwar ds but confirms its so uther n d istribution. Ho wever,Ma rtina a n d Stewart (1982: Can. J . Earth Sci., 19, 2, 324-327) assigned to Ichthyo r ­nithes a single centrum of ve rtebra com in g from t he Ca mpanian Verm ilion R iverF ormation, Pembina Member, a rea of Morden, Manitoba (Canada); if cor r ectlyr ecogn ized , t his vert eb ra w ould ex ten d t he range of Ichthy ornithes far northward.Martin and Stewart also decl a r ed t he r ecord of I ch t hyor n is from the Aust in Chalk,T exas (as r eferred a bove) to be er ro neo us.

Terrestrial birdsMourer- Chauvire (1981: Geob ios 14, 5, 637-647, an d 1982: Geobios 15, 2, 268- 269)

came t o the conclusion t hat the Phor usrhac idae mus t have di fferentiated as ea r ly asthe Cret aceou s. They originate most prob ably from South America where they areknown since the Oligocene a nd Moure-Chauv ire has recently d iscovered one of them,Ameghinornis m in or , in t he Oligocene of Quercy, France. Now, there are only twopossib le di sp ersal r ou t es t o Europe: e ither t hrou gh North America or vi a Afr ica.Either of these two continents los t any con nection t o South America before thebegin n ing of Tertiary and therefore the dispersal must have occured during theCret aceous. The same applies t o the Idiornithidae an d Bathornithidae which a ll cer­tainly appear ed wit h an old r ad iat ion of the ra ll iforms.