A New Presbyornithid Bird (Aves, Anseriformes) from the Late Cretaceous of Southern Mongolia

Home , Presbyornis, Presbyornithidae

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3386, 11 pp., 3 ®gures, 1 table December 27, 2002

EVGENY N. KUROCHKIN,1,3 GARETH J. DYKE,2 AND ALEXANDR A. KARHU1

ABSTRACT We describe a new large representative of the important fossil anseriform taxon Presbyor- nithidae from the latest Cretaceous (Maastrichtian) Nemegt Formation of southern Mongolia. This new taxon, Teviornis gobiensis, n. gen. et n. sp., is known from the associated manual portion of a right wing and the distal end of a right humerus, but is clearly diagnosable with respect to all other known representatives of the fossil Presbyornithidae. It is placed within the clades Anseriformes and Presbyornithidae, respectively, on the basis of a number of de- rived characters of the carpometacarpus and digits. Importantly, description of Teviornis con- ®rms the presence of members of the neornithine clade Anseriformes (``waterfowl'') in the Late Cretaceous, as has been suggested previously on the basis of much less diagnostic fossil material as well as from clade divergence estimates founded on molecular sequence data. The extinct Presbyornithidae thus has a worldwide distribution and ranged in age from at least the Maastrichtian through to the uppermost Eocene.

INTRODUCTION (Olson, 1985; Unwin, 1993). A great deal of controversy, however, surrounds the question The fossil remains of modern birds (ϭ of the timing and extent of the diversi®cation Neornithes sensu Cracraft, 1988) are abun- of these so called ``modern'' clades by the dant and well represented in early Tertiary time of the Cretaceous-Tertiary (K-T) bound- deposits from all over the world, but partic- ary. Although some interpretations of the ularly in Asia, Europe, and North America fossil record have led to the proposal that the

1 Paleoherpetology and Paleornithology Laboratory, Paleontological Institute of the Russian Academy of Sciences, 123 Profsouznaya Street, Moscow GSP-7, 117997, Russia. e-mail: [email protected] 2 Chapman Research Fellow, Division of Vertebrate Zoology (Ornithology), American Museum of Natural History. Present address: Department of Zoology, University College Dublin, Bel®eld Dublin 4, Ireland. e-mail: gareth. [email protected] 3 To whom correspondence should be addressed.

Copyright ᭧ American Museum of Natural History 2002 ISSN 0003-0082 2 AMERICAN MUSEUM NOVITATES NO. 3386 bulk of the neornithine diversi®cation oc- Paleontological Expedition4. This new taxon curred in the aftermath of the K-T event is similar in size to both the known speci- (Feduccia, 1999), recent ``molecular clock'' mens of Presbyornis isoni Olson (from the hypotheses have estimated some divergences Upper Paleocene of Maryland, USA; Olson, to have been deep in the Cretaceous (e.g., 1994) and to the less well-known Headon- Cooper and Penny, 1997; Rodionov, 1997; ornis hantoniensis Harrison and Walker Van Tuinen et al., 2000; Paton et al., 2002). (from the Upper Eocene of England; see also As a result, recent debates have focused on Dyke, 2001b). However, comparisons with the question of just how deep the extant lin- both these larger taxa as well as the smaller eages of birds can be traced prior to the K- Presbyornis pervetus Wetmore, Telmabates T boundary (Kurochkin, 1995; Cracraft, antiquus Howard, and large numbers of ad- 2001) in the absence of well-dated and even ditional bones referred to Presbyornithidae partially complete fossil material from the from the Mongolian Lower Eocene locality of Tsagaan Khushuu5 present a number of Mesozoic (Dyke and Mayr, 1999; Dyke, osteological features that allow us to diag- 2001a). nose an additional taxon. In spite of the known diversity of birds The new specimen (PIN 4499-1) is from from the Paleocene and Eocene (e.g., Brod- the locality of Gurilyn Tsav in the central korb, 1964; Olson, 1985; Unwin, 1993; MlõÂ- part of the Umnogobi Aimak, east of the Al- kovsky, 2002), thus far the well-known fossil tan Ula Ridge, but close to the well-known anseriform Presbyornis Wetmore remains the fossil site of Bugeen Tsav (®g. 1). Sediments most abundant described taxon of this age at Gurilyn Tsav form the lower portions of simply in terms of the sheer numbers of the Nemegt Horizon, at the base of the the specimens that have been recorded (Olson Nemegt Formation, and have been dated and Feduccia, 1980; Olson, 1985; Ericson, Campanian to Early Maastrichtian in age 2000). Presbyornis, along with the other (Martinson, 1982; Barsbold, 1983; Shuvalov, members of the clade Presbyornithidae, has 2000; Khand et al., 2000). Fossil vertebrates been described from Tertiary localities in are relatively common from Gurilyn Tsav; both northern and southern hemispheres taxa that have been collected (some of which (Wetmore, 1926; Howard, 1955; Harrison remain undescribed) include a variety of the- and Walker, 1978; Olson, 1985; Ericson, ropods and hadrosaurian dinosaurs, as well 1997, 2000; Dyke, 2001b) as well as puta- as varanid lizards, turtles, ®sh, and the large tively from the latest stages of the Mesozoic enantiornithine bird Gurilynia (see Kuroch- (Noriega and Tambussi, 1995; Ericson, 2000; kin, 1999). Hope, 2002; Hope and Stidham, 2001). For many years, Presbyornis and its kin However, many Cretaceous specimens thus were considered intermediate in their mor- far considered are poorly constrained tem- phology, apparently combining characters porally and preserve too few anatomical characters to be informative within a phylo- 4 Although PIN 4499-1 was collected in 1987, its identi®cation as presbyornithid was not seriously con- genetic context (Clarke and Chiappe, 2001; sidered at ®rst because of both its Cretaceous age and Chiappe and Dyke, 2002). The recognition the fact that all other records for these birds were from of well-dated Cretaceous fossil material that the Paleocene and Eocene. However, at a round table can be demonstrated to have certain af®nities discussion at the Fourth Meeting of the Society of Avian Paleontology and Evolution in Washington, D.C., in within Neornithes is imperative to further ad- June 1996 (Olson et al., 1999), data were presented that dress questions relating to both the timing documented the presence of presbyornithids from the and extent of the evolutionary radiation of Cretaceous(?) of Antarctica (Noriega and Tambussi, 1995). Notes from members of the original 1987 PIN modern birds before, and after, the K-T ®eld team indicated that more remains of this single boundary (Dyke, 2001a; Kurochkin, 2001). specimen were initially collected from this locality. This We describe a fossil presbyornithid bird material was subsequently damaged and lost. that was collected in 1987 from the Late Cre- 5 Olson (1985: 171) and Ericson (2000: 2) have referred to our data regarding the presence of Presbyornis taceous of southern Mongolia by an expedi- sp. in the Paleocene of Mongolia (Tsagaan Khushuu lo- tionary team of the Joint Russian-Mongolian cality). 2002 KUROCHKIN ET AL.: PRESBYORNITHID BIRD 3 seen in both extant ducks and ¯amingos. Thus members of Presbyornithidae were widely cited as providing evidence for an evolutionary link between a number of the traditional avian orders by the time of the earliest Eocene (Olson and Feduccia, 1980; Feduccia, 1995, 1999). However, the recent inclusion of Presbyornis within numerical analyses of Recent waterfowl have demonstrated that this taxon can be placed well within order Anseriformes (commonly referred to as ``waterfowl''; also comprising Fig. 1. Map showing the Mongolian Late Cre- the extant screamers, ducks, and geese), taceous localities of Gurilyn Tsav (1), where PIN closely related to the true ducks (Anatidae; 4499-1 was found, and Bugeen Tsav (2). Both see Ericson, 1997, 2000; Livezey, 1997, have been dated Campanian to Early Maastrich- 1998; see also Paton et al., 2002). tian in age (see text for details).

MATERIALS AND METHODS golia (this paper) as well as putatively from Our use of anatomical terminology fol- Antarctica and North America (Noriega and lows Baumel and Witmer (1993), with some Tambussi, 1995; Hope and Stidham, 2001; modi®cations to English after Howard see Discussion), the Upper Paleocene±Lower (1929). For discussion of diagnosis and char- Eocene of North America (Wetmore, 1926; acter distributions within Anseriformes we Olson and Feduccia, 1980; see also Ericson, follow Livezey (1997, 1998) and Ericson 2000), Argentina (Howard, 1955), and Mon- (1997); within fossil Presbyornithidae we golia (Kurochkin, 1988), as well as the Mid- follow Ericson (2000). Institutional abbrevi- dle±Upper Eocene of England (Harrison and ations are as follows: AMNH, American Walker, 1978; see also Dyke, 2001b). Isolat- Museum of Natural History, New York; ed elements that may also be presbyornithid BMNH PAL, The Natural History Museum, have been reported from the Late Paleocene London (Palaeontology Department Collec- of Kazakhstan (Nessov, 1988), and the Eo- tions); MLP, Department of Vertebrate Pa- cene of Khirghizia (personal commun. cited leontology, Museo de la Plata; PIN, Pale- in Ericson, 2000). ontological Institute of the Russian Academy of Sciences, Moscow; USNM, United States National Museum, Washington D.C. Teviornis gobiensis, new genus and species Figures 2, 3, table 1

SYSTEMATIC PALEONTOLOGY ETYMOLOGY: For Victor Tereschenko, pa- AVES LINNAEUS, 1758 leontologist at the PIN, who found the spec- ANSERIFORMES (``WATERFOWL'') imen; ornis (Greek), bird (masculine); Gobi, WAGLER, 1831 a severe desert in Central Asia. ANATOIDEA (``TRUE WATERFOWL'') SENSU HOLOTYPE: PIN 44991-1, partially crushed LIVEZEY, 1997 PRESBYORNITHIDAE WETMORE, 1926 associated remains of a right forelimb, in- cluding a fragment of the distal humerus, TYPE GENUS: Presbyornis Wetmore, 1926. both the radiale and ulnare, the complete car- CURRENTLY INCLUDED GENERA: Presbyor- pometacarpus, phalanx 1 of digiti majoris, nis Wetmore, 1926; Telmabates Howard, and the phalanx of digiti minoris (®gs. 2, 3). 1955; Headonornis Harrison and Walker, TYPE LOCALITY AND HORIZON: Gurilyn 1976 (Harrison and Walker, 1979); Teviornis, Tsav, Nemegt Basin (®g. 1); 15 km north of n. gen. (this paper). the Altan Ula ridge and 7 km east of Bugeen STRATIGRAPHIC AND GEOGRAPHIC DISTRI- Tsav; north west corner of Umnogobi Aimak, BUTION: Members of Presbyornithidae are Mongolia; Nemegt Formation, lower Nemegt known from the Upper Cretaceous of Mon- Horizon, Upper Cretaceous (Maastrichtian). 4 AMERICAN MUSEUM NOVITATES NO. 3386

Fig. 2. Teviornis gobiensis, new genus and species, holotype PIN 4499-1. A, Right carpometacarpus in dorsal view; B, Right phalanx 1 of digiti majoris in dorsal view; C, Right phalanx of digiti minoris in dorsocaudal view; D, Right carpometacarpus in ventral view; E, Distal end of the right carpometacarpus in distal view; F, Proximal end of right phalanx 1 of digiti majoris in proximal view; G, Fragment of the right humerus in dorsal view; H, in cranial view; and I, distal view. Specimens have been coated with ammonium chloride to enhance contrast. Scale bar is 1 cm.

KNOWN DISTRIBUTION: Upper Cretaceous of trochlea carpalis well developed and con- (Maastrichtian) of southern Mongolia. nected to the dorsal emargination of os me- DIAGNOSIS: Our referral of Teviornis to tacarpale majus (see also Howard, 1955); Anseriformes is based, in the ®rst place, on presence of well-developed scars for lig. ul- clear similarity with the other known mem- nocarpometacarpale dorsale and lig. radi- bers of Presbyornithidae. In addition, the pre- ocarpometacarpale dorsale; presence of a served carpometacarpus of this new taxon small canalis interosseus distalis in fossa in- exhibits the following character listed by fratrochlearis. Livezey (1997) as diagnostic for Anserifor- DIFFERENTIAL DIAGNOSIS: Teviornis is dis- mes (based on cladistic analysis to the exclu- tinguished from the other known members of sion of Tinamidae and Galliformes; Livezey, Presbyornithidae on the basis of the follow- 1998): absence of craniocaudal curvature of ing character conditions: tuberculum supra- corpus of carpometacarpus (os metacarpale condylare dorsale of humerus extended far minus) relative to os metacarpale majus. Tev- proximally; fossa infratrochlearis of carpo- iornis is placed within Presbyornithidae on metacarpus stretched markedly craniocaudal- the basis of the following character condi- ly; proximal portion of the branch of os me- tions as listed by Ericson (2000): dorsal ridge tacarpale minus widened both dorsoventrally 2002 KUROCHKIN ET AL.: PRESBYORNITHID BIRD 5

Fig. 3. Drawings of the preserved elements of the right wing of Teviornis gobiensis, new genus and species, holotype PIN 4499-1. Carpometacarpus in dorsal (A), ventral (D), proximal (E), cranial (F), and distal (G) views. Phalanx 1 of digiti majoris in dorsal (B), ventral (J), and proximal (H) views. Phalanx of digiti minoris in dorsocaudal view (C) and in ventrocranial view (I). Ulnare in dorsal (K) and caudal (L) views. Radiale in cranial (M) and caudal (N) views. Fragment of distal end of the humerus in cranial (O), dorsal (P), and distal (Q) views. Abbreviations: ads, area distal of sulcus tendineus; adm, facies articularis digiti minoris; dmm, distal branch of os metacarpale majus; cpe, cranial end of processus extensorius; cpp, area cranial of proc. pisiformis; csp, cranial side of the phalanx 1; ea, excavated area between condylus dorsalis and tuberculum supracondylare dorsale; ed, epicondylus dorsalis; ®t, fossa infratrochlearis; pp, processus pisiformis; prm, proximal area of os metacarpale minus; saf, smaller portion of facies articularis digiti majoris; ss, sulcus scapulotricipitalis; ts, tuberculum supracondylare dorsale; vaf, ventral metacarpal articular facet. Scale bar is 1 cm. and craniocaudally with respect to its distal donornis; hence, further differentiation with continuation; facies articularis digitis minoris respect to this taxon must await the recovery on os metacarpale minus divided by a cran- of additional fossil material. On the basis of iocaudal groove into two distinct facets; dis- the carpometacarpus (as far as this element tal extension of facies articularis digitalis ma- is known in the genera of Presbyornithidae), jor widened distally; ventral portion of facies we consider Teviornis to be a distinct taxon articularis digitalis major elevated in the di- (see below). rection of the tubercle for the insertion of m. TYPE SPECIES: Teviornis gobiensis; only abductor digiti majoris. It should be noted, currently recognized species. however, that although portions of the hu- MEASUREMENTS (in mm): Craniocaudal merus are known for Presbyornis isoni (Ol- width of condylus ventralis of the humerus, son, 1994), Dyke (2001b) only tentatively re- 5.3; length of carpometacarpus, 65; dorso- ferred this element to the Eocene genus Hea- ventral height of proximal end of carpome- 6 AMERICAN MUSEUM NOVITATES NO. 3386

TABLE 1 Selected Measurements of Preserved Wing Elements for Members of Presbyornithidae (from various sources)

V. gobiensis P. pervetus P. recurvirostris P. isoni T. antiquus Carpometacarpus Greatest length 65 44±54 63 Proximal width 15.2 9.7±11.9 13 15.8±16.6 Distal width 7.7 5.1±7.1 7.1±7.4 Phalanx 1 of major digit Greatest length 30.8 19.8±27.6 40.6 29.7±30.7 Proximal width 8.3 4.3±5.8 9.5 tacarpus, 7.2; craniocaudal width of proximal listed by Ericson (2000) as characteristic for end of carpometacarpus across the processus Presbyornithidae are lacking in PIN 4499-1, extensorius, 15.2; dorsoventral height of dis- with the exception of a weakly developed tal end of carpometacarpus, 7.7; length of sulcus scapulotricipitalis (resulting from the phalanx 1 digiti majoris, 30.8; ventral prox- small epicondylus dorsalis in this specimen). imal width of phalanx digiti majoris, 8.3; ULNARE: The small but remarkably well- length of phalanx digiti minoris, 13.9; prox- preserved ulnare of Teviornis (®g. 3K, L) ex- imal width of phalanx digiti minoris, 4.2 (for hibits two characters listed by Ericson (2000) comparative measurements, see table 1). as distinctive to Presbyornithidae, a distinct but shallow sulcus on its lateral side as well DESCRIPTION AND COMPARISONS as a wide incisura metacarpalis. These characters, however, are based entirely on the HUMERUS (®gs. 2G±I, 3O±Q): The portion of the distal right humerus that is preserved known material of Presbyornis pervetus, the as part of PIN 4499-1 is broken ventral to only presbyornithid other than Teviornis for the condylus ventralis and across the remain- which this element is known. der of the fossa m. brachialis. However, the RADIALE: The radiale is also represented as shapes of both the dorsal and ventral con- part of PIN 4499-1 (®g. 3M, N). As noted dyles are discernible; notably, a well-exca- by Ericson (2000) again for Presbyornis vated depression is present between the tu- pervetus, this element has a mediolaterally berculum supracondylare dorsale and the compressed caudal end, a deep sulcus on its condylus dorsalis that is not evident in spec- proximal side with a pronounced and cen- imens referred to either Presbyornis or Tel- trally located tuberculum cranial to it. The mabates (Olson, 1994; Ericson, 2000). Since incisura on the cranial edge of the radiale in a depression of similar depth and width is Teviornis is also wide and shallow. also seen in BMNH PAL A 5105, a distal CARPOMETACARPUS: The best preserved end of a left humerus referred to the large and most anatomically informative element Headonornis by Harrison and Walker (1979; of PIN 4499-1 is an almost complete right also ®gured by Dyke, 2001b), but it is not carpometacarpus (®gs. 2A±D, 3A±H). Be- evident on the holotype humerus of the sim- cause this element is well known in the other ilarly sized Presbyornis isoni (USNM presbyornithids with the exception of Pres- 294116; Olson, 1994), this feature does not byornis isoni (Olson, 1994) and Headonornis seem to be related to overall size (see Dis- (Dyke, 2001b), it forms the basis of our di- cussion). Teviornis is also distinguished from agnosis for Teviornis. As is the case in the Presbyornis and Telmabates because the tu- known presbyornithids (Ericson, 2000), the berculum supracondylare dorsale is remark- Lower Eocene Anatalavis oxfordi Olson (Ol- ably far extended proximally in the former. son, 1999), and the basal Anseriformes An- Because of the poor preservation of this el- hima, Chauna (Anhimidae), and Anseranas, ement, most of the morphological features the distal end of the carpometacarpus bears 2002 KUROCHKIN ET AL.: PRESBYORNITHID BIRD 7 a well-developed and deep sulcus tendineus PIN 4499-1 is stretched dorsoventrally on its dorsal surface that extends at least one- whereas it is much shorter in both Telma- third of its length. Although only the proxi- bates and Presbyornis pervetus (perhaps as a mal and distal ends of the right carpometa- result of the much larger size of Teviornis). carpus and the distal end of the left are pre- The articular facet for phalanx digiti minoris served as part of the holotype of Telmabates is undivided in the two smaller taxa but is (AMNH 3170), this conformation of the sul- grooved craniocaudally in Teviornis, and the cus is seen in a number of the additional area distal to the sulcus tendineus (in dorsal specimens referred to this taxon (Howard, view) bears no depression (in both Telma- 1955). bates and Presbyornis pervetus this area is On the proximal end of the carpometacar- excavated to varying degrees). pus, the fossa infratrochlearis of PIN 4499-1 Ericson (2000) listed a number of char- is stretched craniocaudally to a greater extent acters of the carpometacarpus as diagnostic than it is in either Presbyornis pervetus or for Presbyornithidae, the majority of which Telmabates, which both have a much more are seen clearly in PIN 4499-1. Of these fea- circular contour of this fossa. The angle be- tures, the following are preserved in the ho- tween the processus extensorius and the lotype of Teviornis: carpometacarpus straight trochlea carpalis is almost perpendicular, as with major and minor metacarpals parallel; opposed to being distinctly curved, as is the caudal portion of dorsal part of trochlea car- case in the other presbyornithids. In this re- palis well developed and connected to the spect, this element of Teviornis closely re- edge of the major metacarpal; presence of sembles another large and as yet undescribed large scars for the insertions of lig. ulnocar- presbyornithid carpometacarpus from the pometacarpale dorsale and lig. radiocarpo- Lower Eocene London Clay Formation of metacarpale dorsale; small vascular foramen England (BMNH PAL A 6241). present in the fossa infratrochlearis; marked The proximal portion of the os metacar- and deep sulcus tendineus; sulcus interosseus pale minus in Teviornis is distinctly widened deep, having an ossi®ed bridge. dorsoventrally when viewed in dorsal view PHALANX 1DIGITI MAJORIS: The proximal as opposed to being equal in width, as is the phalanx of the major digit (®gs. 2B, 3B, J) case in both Presbyornis pervetus and Tel- in Teviornis is long and slender and bears mabates. Just as in Telmabates, the dorsal two deep depressions on its dorsal side. On part of the trochlea carpalis in PIN 4499-1 the proximal side of this element there is a extends as far caudally as does the ventral blunt and cranioventrally protruding tuber- part of the trochlea carpalis; in Presbyornis culum, no dorsal tendinal sulcus on the pila pervetus, the dorsal ridge of the trochlea car- cranialis, and no distal projection on the cau- palis extends far caudally with respect to the dal margin. All three of these characters are ventral ridge of the trochlea carpalis. In both seen in members of Presbyornithidae; this el- Teviornis and Presbyornis the processus pis- ement is known for both Presbyornis (Olson, iformis lies on the same level as os metacar- 1994; Ericson, 2000) and Telmabates pale majus, but in Telmabates this process (AMNH 3180; Howard, 1955). The pre- follows the line of extension of os metacar- served phalanx 1 of Teviornis exhibits a fur- pale minus. Variation in the relative degree ther two characters that allow taxonomic dif- of excavation of the area cranial to the pro- ferentiation from other taxa: a straight cranial cessus pisiformis is seen between the three margin (in dorsal view) and an elevated ven- genera: in Victornis this area is somewhat ex- tral portion of the metacarpal articular facet cavated, whereas it is shallow in Telmabates (in the direction of the tubercle for the in- and very deep in Presbyornis pervetus. Fur- sertion of m. abductor digiti majoris). In oth- ther, the fossa infratrochlearis in PIN 4499-1 er presbyornithids, this phalanx is not raised bears several pronounced muscle scars and ventrally and has a curved cranial margin excavations (as is seen in Presbyornis)Ðin when viewed dorsally (e.g., AMNH 3180 Telmabates this fossa is almost ¯at (Howard, [Telmabates], USNM 294117 [P. isoni], 1955). USNM 483164 [P. pervetus]). The distal end of the carpometacarpus of PHALANX DIGITI MINORIS: The small, well- 8 AMERICAN MUSEUM NOVITATES NO. 3386 preserved phalanx of the minor digit of Tev- of Antarctica). It is however of note that re- iornis (®g. 2C, I) is ¯attened dorsoventrally cords of ``Presbyornis-like'' birds from the and has a sharp and projected pila caudalis Mongolian Cretaceous have been listed pre- and a well-developed tuberculum located dis- viously by some authors (Olson, 1985; Ku- tally with respect to the pila caudalis. These rochkin, 1995) on the basis of a small, dam- features were described for Presbyornis aged, and isolated tarsometatarsus ®gured as pervetus by Ericson (2000). Presbyornithidae(?) by Kurochkin (1995). This single specimen, from the Udan Sayr DISCUSSION locality within the Baruungoyot Formation, is considered Aves incertae sedis here as it Previous studies have demonstrated that preserves no anatomical features that can be the Presbyornithidae were an extremely considered diagnostic of either Presbyorni- widespread group of birds during early Ter- thidae or Anseriformes (Ericson, 2000). tiary times. Members of this clade have been The internal composition of Presbyorni- recovered from the Paleocene±Eocene of Eu- thidae remains controversial and complex. rope (Harrison and Walker, 1978, 1979; For more than 100 years the remains of fossil Dyke, 2001b), South America (Howard, birds bearing varying degrees of resemblance 1955; Cracraft, 1970), and Asia. Indeed, to living Anseriformes have been described these birds are probably the most abundantly from the earliest Tertiary (e.g., Owen, 1846; known taxa from throughout the Tertiary of Milne-Edwards, 1867±1871; Lydekker, North America (see, for example, McGrew 1891; Kurochkin, 1976), but only recently and Feduccia, 1974; Olson and Feduccia, have clear character diagnoses for this group 1980; Olson, 1994; Leggitt and Bucheim, been presented (Livezey, 1997, 1998; Eric- 1997; Benson, 1999; Feduccia, 1999; Eric- son, 1999, 2000). The genus Telmabates, for son, 2000). Recent phylogenetic studies have example, originally described by Howard placed Presbyornis within Anseriformes, (1955) from the Eocene of Patagonia has var- close to the divergence of the Anatidae, the iously been considered to be synonymous true ducks, geese, and swans (Ericson, 1997; with Presbyornis, especially following the Livezey, 1997, 1998), indicating that the di- description of the giant Presbyornis isoni by vergence of this large extant taxon must have Olson (1994). Olson (1985, 1994), for ex- occurred earlier than the oldest known mem- ample, has considered that Presbyornithidae ber of Presbyornithidae. Our description of comprises simply a number of differently Teviornis serves to con®rm previous sugges- sized species within the single genus Pres- tions that Anseriformes did indeed occur pri- byornis. Recently, Ericson (2000: 7, 11) has or to the K-T boundaryÐexisting proposals shown that not only do the two genera vary that this was the case have been based either signi®cantly in the relative sizes of their on more incomplete fossil material (i.e., sin- hindlimb bones (those of Telmabates are also gle and isolated elements presenting few somewhat more robust; cf. AMNH 25552, characters for comparisons; e.g., Hope and 28505 [slabs containing Presbyornis perve- Stidham, 2001; Hope, 2002) or on records tus remains] and AMNH 3167, 3169, 3180, collected from rocks that are of uncertain 3171 [referred specimens of Telmabates]), Cretaceous age (e.g., Olson and Parris, 1987; but that there are clear osteological differ- Noriega and Tambussi, 1995; see contrary ci- ences between them (e.g., linea muscularis tations in Olson, 1994; Clarke and Chiappe, cranialis of coracoid obsolete in Presbyornis 2001). Comparisons between PIN 4499-1 [welldeveloped in Telmabates.] More prob- and other putative records of Presbyornithi- lematical with respect to these diagnoses are dae from the ?Cretaceous are not possible the less completely known taxa Headonornis here, either because of the fragmentary pres- hantoniensis Harrison and Walker and Pres- ervation of the available material (see illus- byornis isoni Olson. Although the holotype trations in Hope, 2002) or because of the lack of Headonornis is an isolated coracoid of corresponding skeletal elements (as is the (BMNH PAL A 30325), Dyke (2001b) case in MLP 93-I-3-1 described by Noriega showed that incomplete portions of humeri and Tambussi [1995] from the ?Cretaceous (BMNH PAL A 3686, A 5105) referred to 2002 KUROCHKIN ET AL.: PRESBYORNITHID BIRD 9 this taxon by Harrison and Walker (1976, to specimens in their care, and Luis Chiappe 1979) are in fact referrable to P. isoni (on and Gerald Mayr for their helpful comments the basis of a more recently collected and on the manuscript. This research was sup- more complete humerus, BMNH PAL A ported by grants from the Russian Founda- 6240). It remains entirely possible that these tion for Basic Research no. 00-04-49348 and two large presbyornithids, P. isoni and Hea- No. 00-15-97754 and the Frank. M. Chap- donornis, are actually the same taxon, but the man Fund of the Department of Ornithology, discovery of additional material will be re- American Museum of Natural History. quired to con®rm this (Dyke, 2001b). Be- cause Presbyornis isoni is currently known REFERENCES largely on the basis of the humerus (Olson, 1994; Benson, 1999), only a few of the dif- Barsbold, R. 1983. Carnivorous dinosaurs from ferential characters listed by Ericson (2000) the Cretaceous of Mongolia. Trudy Sovmestnoi (with respect to Telmabates) apply (i.e., im- Sovetsko-Mongol'skoi Paleontologicheskoi Ek- pressio m. pronator super®cialis on distal hu- speditsii 19: 1±120. [in Russian] merus extends much farther than lig. colla- Baumel, J.J., and L.M. Witmer. 1993. Osteologi- ca. In J.J. Baumel, A.S. King, J.E. Breazile, terale ventrale in Presbyornis). An additional H.E. Evans, and J.C. Vanden Berge (editors), specimen, BMNH PAL A 43164, an incom- Handbook of avian anatomy: nomina anatomi- plete portion of sternum described by Lydek- ca avium: 45±132. Publications of the Nuttall ker (1891) as the holotype of Proherodius Ornithological Club 23. oweni, is very similar in its size and overall Benson, R.D. 1999. Presbyornis isoni and other morphology (e.g., in the shape of the cora- late Paleocene birds from North Dakota. Smith- coidal sulci) to this element in P. pervetus sonian Contributions to Paleobiology 89: 253± (Dyke, 2001b); although tantalizingly incom- 259. plete, it remains possible that this specimen Brodkorb, P. 1964. Catalogue of fossil birds, part is in fact the earliest described presbyornithid 2 (Anseriformes through Galliformes). Bulletin of the Florida State Museum Biological Scienc- (as suggested by Harrison and Walker, 1978). es 8: 195±335. Whatever the ®nal outcome of these nig- Chiappe, L.M., and G.J. Dyke. 2002. The Meso- gling taxonomic problems, it is clear that zoic radiation of birds. Annual Reviews of Presbyornithidae was one of the most abun- Ecology and Systematics 33: 91±124. dant groups of Late Cretaceous±Tertiary Clarke, J.A., and L.M. Chiappe. 2001. A new car- neornithine birds. The abundance of these inate bird from the Late Cretaceous of Argen- birds in deposits of this age is probably a tina. American Museum Novitates 3323: 1±23. result of their aquatic or semiaquatic ecolo- Cooper, A., and D. Penny. 1997. Mass survival of gy, resulting in their differential preservation birds across the Cretaceous±Tertiary boundary: across the K-T boundary (when compared to molecular evidence. Science 275: 1109±1113. Cracraft, J. 1970. A new species of Telmabates other groups of largely terrestrial taxa). Since (Phoenicopteriformes) from the Lower Eocene a similar bias is seen in the fossil record of of Patagonia. Condor 72: 479±480. non-neornithine birds throughout the Creta- Cracraft, J. 1988. The major clades of birds. In ceous (Padian and Chiappe, 1998; Chiappe M.J. Benton (editor), The phylogeny and clas- and Dyke, 2002), it seems logical to infer si®cation of the tetrapods, vol. 1: amphibians, that members of other clades of modern birds reptiles, birds: 333±355. Oxford: Systematics were also present prior to the terminal Cre- Association. taceous extinction event; these have either Cracraft, J. 2001. Avian evolution, Gondwana yet to be found or have not so far been rec- biogeography and the Cretaceous-Tertiary mass ognized convincingly on the basis of phylo- extinction event. Proceedings of the Royal So- genetically informative fossil material. ciety of London Series B Biological Sciences 268: 459±469. Dyke, G.J. 2001a. The evolutionary radiation of ACKNOWLEDGMENTS modern birds: systematics and patterns of diversi®cation. Geological Journal 36: 305±315. We thank Sandra Chapman (BMNH), Dyke, G.J. 2001b. The fossil waterfowl (Aves: Chris Collins, Paul Sweet (AMNH), and Anseriformes) from the Eocene of England. Mark Florence (USNM) for providing access American Museum Novitates 3354: 1±15. 10 AMERICAN MUSEUM NOVITATES NO. 3386

Dyke, G.J., and G. Mayr. 1999. Did parrots exist Kurochkin, E.N. 1988. Cretaceous birds of Mon- in the Cretaceous period? Nature 399: 317±318. golia and their signi®cance for the study [of Dyke, G.J., B.E. Gulas, and T.M. Crowe. 2003. the] phylogeny of class Aves. In E.N. Kuro- Suprageneric relationships of galliform birds chkin (editor), Fossil reptiles and birds of Mon- (Aves, Galliformes): a cladistic analysis of golia. Trudy Sovmestnoi Sovetsko-Mongol'skoi morphological characters. Zoological Journal Paleontologicheskoi Ekspeditsii: 33±42. Mos- of the Linnean Society, in press. cow: Nauka. [in Russian] Ericson, P.G.P. 1997. Systematic position of the Kurochkin, E.N. 1995. Synopsis of Mesozoic paleogene family Presbyornithidae (Aves). birds and early evolution of class Aves. Ar- Zoological Journal of the Linnean Society 121: chaeopteryx 13: 47±66. 429±483. Kurochkin, E.N. 1999. A new large enantiornithid Ericson, P.G.P. 1999. New material of Juncitarsus from the Late Cretaceous of Mongolia. In I. (Phoenicopteriformes), with a guide for differ- Darevskii and A. Averianov (editors), Materials entiating that genus from the Presbyornithidae on the history of fauna of Eurasia. Zoological (Anseriformes). Smithsonian Contributions to Institute of the Russian Academy of Sciences, Paleobiology 89: 245±251. Saint-Petersburg 277: 132±147. [in Russian] Ericson, P.G.P. 2000. Systematic revision, skeletal Kurochkin, E.N. 2001. New ideas about the origin anatomy, and paleoecology of the New World and early evolution of birds. In E.N Kurochkin early Tertiary Presbyornithidae (Aves: Anseri- and I. Rakchimov (editors), Achievements and formes). Paleobios 20: 1±23. problems of ornithology of northern Eurasia on Feduccia, A. 1995. Explosive evolution in Tertia- a boundary of centuries: 68±96. Kazan: Ma- ry birds and mammals. Science 267: 637±638. garif Publishers. [in Russian with English sum- Feduccia, A. 1999. The origin and evolution of mary] birds, 2nd ed. New Haven, CT: Yale University Leggitt, V.L., and H.P. Buchheim. 1997. Pres- Press. byornis (Aves: Anseriformes) eggshell from Harrison, C.J.O., and C.A. Walker. 1976. Birds of three avian mass mortality sites: Eocene Fossil the British Upper Eocene. Zoological Journal Lake, Lincoln County, Wyoming. Journal of of the Linnean Society 59: 323±351. Vertebrate Paleontology 17: 60A. Harrison, C.J.O., and C.A. Walker. 1978. Pro- Livezey, B.C. 1997. A phylogenetic analysis of herodius oweni as a species of Presbyornithi- basal Anseriformes, the fossil Presbyornis and dae. Tertiary Research 2: 1±3. the interordinal relationships of waterfowl. Harrison, C.J.O., and C.A. Walker. 1979. Birds of Zoological Journal of the Linnean Society 121: the British Lower Oligocene. Tertiary Research 361±428. Special Paper 5: 29±39. Livezey, B.C. 1998. ErratumÐa phylogenetic Hope, S. 2002. The Mesozoic fossil record of analysis of basal Anseriformes, the fossil Pres- Neornithes. In L.M. Chiappe and L.D. Witmer byornis and the interordinal relationships of (editors), Mesozoic birds: above the heads of waterfowl. Zoological Journal of the Linnean the dinosaurs. Berkeley: University of Califor- Society 124: 397±398. nia Press, in press. Lydekker, R. 1891. Catalogue of fossil birds in Hope, S., and T.A. Stidham. 2001. Late Creta- the British Museum (Natural History). London: ceous-Early Tertiary avifaunas in North Amer- Longmans, 368 pp. ica. Journal of Vertebrate Paleontology 21: Martinson, G.G. 1982. General problems of pa- 63A. laeolimnological studies in Mongolia. In G.G. Howard, H. 1929. The avifauna of Emeryville Martinson (editor), Mesozoic lake basins of Shellmound. University of California Publica- Mongolia: 5±17. Leningrad: Nauka. [in Rus- tions in Zoology 32: 301±394. sian] Howard, H. 1955. A new wading bird from the McGrew, P.O., and A. Feduccia. 1974. A prelim- Eocene of Patagonia. American Museum Nov- inary report on a nesting colony of Eocene itates 1710: 1±25. birds. Wyoming Geological Association Guide- Khand, Yo., D. Badamgarav, Ya Ariunchimeg, book 1973: 163±164. and R. Barsbold. 2000. Cretaceous system in Milne-Edwards, A. 1867±1871. Recherches ana- Mongolia and its depositional environments. In tomiques et paleÂontologiques pour servir aÁ H. Okada and N.J. Mateer (editors), Cretaceous l'histoire des oiseaux fossiles de la France (4 environments of Asia: 49±79. New York: El- volumes). Paris: Victor Masson. sevier Science. MlõÂkovsky, J. 2002. Cenozoic birds of the world, Kurochkin, E.N. 1976. Survey of Paleogene birds part 1: Europe. Praha: Ninox Press. of Asia. Smithsonian Contributions to Paleo- Nessov, L.A. 1988. New Cretaceous and Paleo- biology 27: 75±86. gene birds of Soviet Middle Asia and Kazachs- 2002 KUROCHKIN ET AL.: PRESBYORNITHID BIRD 11

tan and environments. Proceedings of the Zoo- tion. Smithsonian Contributions to Paleobiolo- logical Institute of USSR Academy of Sciences gy 89. 182: 116±123. [in Russian] Owen, R. 1846. A history of British fossil mam- Noriega, J.I., and C.P. Tambussi. 1995. A Late mals and birds. London: John Van Voorst. Cretaceous Presbyornithidae (Aves: Anserifor- Padian, K., and L.M. Chiappe. 1998. The origin mes) from Vega Island, Antarctic peninsula: pa- and early evolution of birds. Biological Re- leobiogeographic implications. Ameghiniana views of the Cambridge Philosophical Society 32: 57±61. 73: 1±42. Olson, S.L. 1985. The fossil record of birds. In Paton, T., O. Haddrath, and A.J. Baker. 2002. D.S. Farner, J.R. King, and K.C. Parkes (edi- Complete mitochondrial DNA genome se- tors), Avian biology vol. 8: 79±256. New York: quences show that modern birds are not de- Academic Press. scended from transitional shorebirds. Proceed- Olson, S.L. 1994. A giant Presbyornis (Aves: An- ings of the Royal Society of London Series B seriformes) from the Paleocene Aquia Forma- Biological Sciences 269: 839±846. tion of Maryland. Proceedings of the Biological Rodionov, A.V. 1997. Evolution of avian chro- Society of Washington 107: 429±435. mosomes and linkage groups. Genetics 33: Olson, S.L. 1999. The anseriform relationships of 725±738. Anatalavis Olson and Parris (Anseranatidae), Shuvalov, V.F. 2000. The Cretaceous stratigraphy with a new species from the Lower Eocene and palaeobiogeography of Mongolia. In M.J. Benton, M.A. Shishkin, D.M. Unwin, and E.N. London Clay. Smithsonian Contributions to Pa- Kurochkin (editors), The age of dinosaurs in leobiology 89: 231±243. Russia and Mongolia: 256±278. Cambridge: Olson, S.L., and A. Feduccia. 1980. Presbyornis Cambridge University Press. and the origin of the Anseriformes (Aves: Unwin, D.M. 1993. Aves. In M.J. Benton (editor), Charadriomorphae). Smithsonian Contributions The fossil record II: 717±737. London: Chap- to Zoology 323: 1±24. man and Hall. Olson, S.L., and D.C. Parris. 1987. The Creta- Van Tuinen, M., C. G. Sibley, and S.B. Hedges. ceous birds of New Jersey. Smithsonian Con- 2000. The early history of modern birds in- tributions to Paleobiology 63: 1±22. ferred from DNA sequences of nuclear and mi- Olson, S.L., P. Wellnhofer, C. Mourer-ChauvireÂ, tochondrial ribosomal genes. Molecular Biolo- D.W. Steadman, and L.D. Martin. 1999. Avian gy and Evolution 17: 451±457. paleontology at the close of the 20th century: Wetmore, A. 1926. Fossil birds from the Green proceedings of the 4th international meeting of River deposits of eastern Utah. Annals of the the Society of Avian Paleontology and Evolu- Carnegie Museum 16: 391±497. Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates and Bulletin published during the last ®ve years are available at World Wide Web site http://library.amnh.org. Or address mail orders to: American Museum of Natural History Library, Central Park West at 79th St., New York, NY 10024. TEL: (212) 769-5545. FAX: (212) 769- 5009. E-MAIL: [email protected]

a This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).