sign in sign up
<<
Home , Penguin

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3423, 18 pp., 9 ®gures, 1 table December 9, 2003

Description of the Earliest from and First Locality of Tierra Del Fuego,

JULIA A. CLARKE,1 EDUARDO B. OLIVERO,2 AND PABLO PUERTA3

ABSTRACT We report the discovery of the ®rst vertebrate from the Paleogene of Tierra del Fuego (Isla Grande), Argentina, in southernmost South America. The specimen consists of parts of an associated pelvic girdle and limb that are identi®ed as belonging to the penguin stem (Aves: Pansphenisciformes). The specimen, from an exposure of the Leticia Formation (late middle ), is the earliest known penguin (pansphenisciform) from South America. It is more than 20 million older than the earliest previously recorded South American pen- guins (from the late ±early ) and, thus, almost doubles their known record on the continent. A detailed description of the new specimen and a discussion of its implications for the understanding of penguin morphological are provided. The new specimen and other fossil do not currently point to the origin of extant, or crown clade, penguin lineages (Spheniscidae), by the Eocene, only to the divergence of the penguin stem lineage from its sister taxon by this time. The new fossil has several morphologies that differ from all extant penguins but are shared with other fossil penguin taxa, suggesting they may be outside Sphen- iscidae. However, in a discussion of the current status of penguin systematics, we suggest the urgent need for comprehensive phylogenetic analysis of fossil and extant penguins to clarify the timing and pattern of penguin diversi®cation. The specimen was recovered from a newly identi®ed fossil vertebrate locality, an exposure of the Eocene Leticia Formation at Punta Torcida on the Atlantic shore of southeastern Tierra del Fuego, Argentina. The new locality is introduced, and a brief geologic description is made,

1 Division of Paleontology, American Museum of Natural History. e-mail: [email protected] 2 Centro-Austral de Investigaciones Cientõ®cas (CADIC) of the Consejo Nacional de Investigaciones Cientõ®cas y TeÂcnicas (CONICET), CC 92, 9410 , Tierra del Fuego, Argentina. 3 Museo PaleontoloÂgico Egidio Feruglio, Avda. Fontana 140, 9100 , Chubut, Argentina.

Copyright ᭧ American Museum of Natural History 2003 ISSN 0003-0082 2 AMERICAN MUSEUM NOVITATES NO. 3423

highlighting the potential of the shallow marine sediments of the Leticia Formation for con- tributing to our knowledge of the Paleogene vertebrate fossil record of Tierra del Fuego, and of southern South America, generally.

INTRODUCTION go, Eocene outcrops are restricted to the western and northern Andean margin and The associated partial penguin pelvic gir- consist mostly of deep marine deposits, with dle and limb described here was recovered the exception of the shallow marine Leticia in the friable, silty, very ®ne sandstones of Formation. Of these exposed deposits, only the Leticia Formation exposed at Punta Tor- the Leticia Formation has produced a fossil cida on the Atlantic coast of southeastern Ti- vertebrate, that which is described here. erra del Fuego, Argentina (®gs. 1±3). The Fossil penguins are known from the late specimen was collected by MarõÂa I. LoÂpez- to late Eocene of , Cabrera and Eduardo B. Olivero during ®eld , which lies off of the northeastern study of the stratigraphy, sedimentology, and tip of the Peninsula. The Eocene paleontology of the Eocene foreland succes- La Meseta Formation exposed on Seymour sion of the Austral Basin. The vertebrate- Island has produced a taxonomically diverse bearing Leticia Formation sandstones repre- penguin assemblage (Wiman, 1905; Simp- sent estuarine and proximal shelf deposits son, 1971a; Cione et al., 1976; Case, 1992). that have also produced abundant gastropods, Most of this material, including at least six bivalves, and solitary corals (Olivero and LoÂ- , is from the upper part of the For- pez-Cabrera, 2001; Olivero et al., 2002). Fo- mation, which is late Eocene in age. This up- raminifera and nannoplankton from the Le- per part is currently included in ``Telm 7'' of ticia Formation indicate a late middle Eocene Sadler (1988), or the Allomember Submeseta age for the deposits (MalumiaÂn et al., 1994; of Marenssi et al. (1998), and has yielded a Olivero and MalumiaÂn, 1999; see below). Sr isotopic age of ϳ34.2 Ma (Dingle and Prior to the recovery of the new specimen, Lavelle, 1998). Sadler (1988) recognized the earliest fossil penguins from South seven informal stratigraphic packages in the America were latest Oligocene or early Mio- La Meseta Formation, ``Telm 1'' to ``Telm cene specimens from (Simpson, 7'', and Marenssi et al. (1998) recognized six 1972; Fordyce and Jones, 1990). At the turn unconformity-bounded allomembers. Less of the 19th century, Carlos Ameghino col- abundant penguin material was also recorded lected penguin material at the mouth of the from ``Telm 3'' and ``Telm 4/5'' or Allo- Santa Cruz River in Santa Cruz Province, members Campamento and Cucullaea, re- Argentina, probably from the Monte LeoÂn spectively. These units are assigned a middle Formation, also of late Oligocene±early Mio- Eocene age based on the ϳ52.4/54.3 (late cene age (Acosta Hospitaleche et al., 2001). early Eocene) Sr isotopic age yielded by the Fossil penguins are additionally known from part of the underlying Acantilados Al- latest Oligocene±early Miocene shallow ma- lomember (Reguero et al., 2002). New dis- rine deposits assigned to the Gaiman or Pa- coveries from Seymour Island include pen- tagonia Formation and exposed in the RõÂo guin bones from the late early Eocene Acan- Chubut valley, near the city of Trelew, Pa- tilados Allomember or Telm 2±3 (La Meseta tagonia (Simpson, 1946; Simpson, 1981; Formation) as well as from the late Paleo- MalumiaÂn et al., 1999 and the bibliography cene Cross Valley Formation (S. Marenssi, therein). personal commun. to E.B. Olivero). Although middle and upper? Eocene ma- Paleogene penguins have been described rine sediments have a regional distribution in from the late middle and late Eocene through southern Patagonia, they outcrop only along Oligocene of (Fordyce, 1991), the western margin of the Patagonian Andes. and further undescribed material is known These deposits are exposed in the Lago Ar- from the early Eocene (Fordyce, personal gentino±RõÂo Turbio area (MalumiaÂn et al., commun.). They are also known from the 1999). In southern and Tierra del Fue- late Eocene (Simpson, 1957; Jenkins, 1974) 2003 CLARKE ET AL.: FOSSIL PENGUIN 3 and possibly Oligocene (Glaessner, 1955) of is also recommended that the name ``Sphen- (reviewed in Fordyce and Jones, iscidae'' be formally applied to the clade 1990). A partial skeleton and other more comprised of the most recent common an- fragmentary material from ``proto-pen- cestor of all extant penguins and all of its guins'', or - propelled divers proposed descendants. However, a formal de®nition of to be closely related to penguins, have ad- Spheniscidae identifying this name as that ditionally been reported from the late Paleo- for the penguin crown clade should await cene or early Eocene of New Zealand (e.g., further consensus on the phylogenetic rela- Fordyce et al., 1986), but have not yet been tionships among extant penguins (e.g., com- described (Fordyce and Jones, 1990; Fordy- pare Kennedy and Page [2002] and clado- ce, personal commun.). gram of O'Hara [1989] published in Wil- liams [1995]) because of the required use of MATERIALS AND METHODS species or specimens as speci®ers for clade names (Cantino and de Queiroz, 2000). In the description of the new specimen, the INSTITUTIONAL ABBREVIATIONS: AMNH, English equivalents of the osteological American Museum of Natural History; nomenclature of Baumel and Witmer (1993) CADIC, Centro Austral de Investigaciones are used. The terms of orientation for the an- Cientõ®cas, Tierra del Fuego, Argentina; atomical position of a , as speci®ed by MoNZ, Museum of New Zealand (formally Clark (1993), were followed with one excep- the Dominion Museum; e.g., as cited in tion. The ``time-honored'' terms (Clark, Simpson, 1971b), Wellington, New Zealand; 1993) of zoological nomenclature ``anterior'' OM, Otago Museum, Dunedin, New Zea- and ``posterior'' were used, rather than ``cra- land. nial'' (and ``rostral'') and ``caudal'' as pro- posed by Clark (1993) in the Handbook of GEOLOGIC SETTING OF THE PUNTA Avian : Nomina Anatomica Avium TORCIDA LOCALITY (Baumel et al., 1993; see further discussion in Clarke and Norell, 2002). Appendix 1 lists In southeastern Tierra del Fuego, the Late the comparative materials used in preparation ±Cenozoic sedimentary ®lling of of the description and systematic discussion. the foreland Austral Basin consists of four All taxonomic names above the species major, unconformity-bounded, synorogenic level used in this paper are used as clade clastic wedges, which are mainly character- names, although most await formal de®nition ized by very thick, deep marine siliciclastic under a system of phylogenetic nomenclature deposits (Olivero and MalumiaÂn, 2002). The (de Queiroz and Gauthier, 1990, 1992) and youngest, Oligocene±Miocene clastic wedge according to the PhyloCode (Cantino and de is mainly exposed outside the folded belt, to Queiroz, 2000). ``Pansphenisciformes'' is the north of Punta Gruesa (®g. 1), and con- used here as a name for all taxa more closely sists of subhorizontal, faulted strata. Within related to extant penguins than to any other the fold and thrust belt, to the south of Punta extant avian taxa. ``Sphenisciformes'' is sug- Gruesa, the deposits of the older clastic gested as a name for all parts of this lineage wedges of ±, late Paleo- with a loss of aerial ¯ight homologous with cene±early Eocene, and late middle to late that of extant penguins. These de®nitions are Eocene±early Oligocene ages, respectively, deliberately not formalized pending recom- reach an aggregate thickness in excess of mendation of the PhyloCode regarding the 3500 m and are highly deformed and mostly proposed use of ``pan'' as a pre®x in all stem exposed in tectonically separated blocks. clade names (Gauthier and de Queiroz, 2001) Nonetheless, superb outcrops along the At- and to allow penguin specialists to debate ap- lantic shore at the Cabo Campo del Medio± propriate de®nitions for these names prior to Punta Torcida anticline (®gs. 1±3) allow for the start date of the PhyloCode (Cantino and the recognition of a very thick (ca. 1600 m) de Queiroz, 2000). Eocene sedimentary succession, comprising As all extant penguins have consistently the Punta Torcida Formation and unnamed been placed in the ``'' Spheniscidae, it strata, early Eocene; the Leticia Formation, 2003 CLARKE ET AL.: FOSSIL PENGUIN 5

Fig. 3. Correlation of sections and sedimentary facies of the lower and upper middle Eocene strata across the Cabo Campo del Medio±Punta Torcida anticline showing the complex geometry and variable thickness of the three main stratigraphic intervals of the Leticia Formation (adapted from Olivero and LoÂpez-Cabrera, 2001). The penguin remains (CADIC P 21; indicated by a star) were recovered near the top of the midshelf sandstone interval of the Leticia Formation near Punta Torcida. dominated by ®ne-grained, glauconitic, es- sandstone bodies, with minor heterolitic, bio- tuarine, and proximal shelf sandstones. These turbated, thin beds, which are interpreted as sandstones show a complex architecture and representing estuarine settings (Olivero and display variable thickness along the northern LoÂpez-Cabrera, 2001; Olivero et al., 2002). and southern margin of the anticline, ca. 200 The scarcity of foraminifera and the absence m and 500 m, respectively (®g. 3), probably of planktonic foraminifera in most of these re¯ecting differential syntectonic subsidence intervals also suggest marginal, shallow, and controlled by a growing fold structure. The restricted marine conditions for the lower Leticia Formation is composed by three main and upper intervals of the Leticia Formation sandstone-dominated intervals. The lower (Olivero and MalumiaÂn, 1999). The middle and upper intervals consist of thick, cross- interval consists of glauconitic, fossiliferous, bedded and parallel-laminated, channeled and highly bioturbated very ®ne sandstones, 6 AMERICAN MUSEUM NOVITATES NO. 3423 representing shelf settings. Minor mudstone 1971b) and Oligocene Archaeospheniscus horizons bear planktonic foraminifera and lowei (OM C 47.27; Simpson 1971b), both nannoplankton indicative of an age con- from New Zealand, but is slightly larger than strained to the G. index Zone to G. incon- the latest Oligocene±early Miocene? Parap- spicua Zone of New Zealand, or equivalent tenodytes antarcticus (e.g., AMNH 3338) to the interval: upper Zone P12 to lower from Patagonia. Zone P14 of tropical areas, that is, late mid- The CADIC P 21 pelvis is represented by dle Eocene (MalumiaÂn et al., 1994; Olivero the left preacetabular iliac blade (®g. 4a, b) and MalumiaÂn, 1999). The penguin bones and a small portion of the left pubis ventro- were recovered near the top of the middle lateral to the acetabulum (®g. 4a). Approxi- shelf sandstone interval, at the southern limb mately one-half of the acetabular rim is pre- of the Cabo Campo del Medio±Punta Torcida served, and no parts of the sacral series or anticline (®g. 3). postacetabular pelvis are preserved. The The overlying Cerro Colorado Formation preacetabular ilium is preserved in ventro- (ca. 855 m thick) consists of a vertical stack- medial view. It was evidently not fused to ing of four coarsening and thickening up- the sacral series; a scar faintly demarcates il- ward successions. Each succession is com- ium contact with the transverse process of a posed of dark gray mudstones at the base, sacral vertebra just anterior to the acetabulum regular intercalation of mudstones and light (®g. 4b). This scar in CADIC P 21 lies in gray or greenish sandstones at the middle approximately the same position as that in part, and thick gray or yellowish ®ne to the extant penguins compared. Just anterior coarse sandstones and pebbly sandstones at to the acetabulum, the ventromedial surface the top. The uppermost succession bears of the blade is slightly concave between the abundant radiolarians and planktonic fora- scar just mentioned and the broken medial minifera typical of an oxygen minimum hab- edge of the blade. itat, indicative of a late Eocene age (upper The ilium in CADIC P 21 broadens to- Zone P15 to upper Zone P16). The micro- ward its anterior end. This end is incomplete, fauna and stratigraphic position of the lower and it is unclear how much of the element is three successions are consistent with a latest missing. The preacetabular portion of the il- middle Eocene age (middle Zone P14 to low- ium in the new fossil may have been rela- er Zone P15, Olivero and MalumiaÂn, 1999). tively short. The shape of the lateral edge of the blade is partially preserved, and the cur- DESCRIPTION OF THE NEW vature of this lateral edge of the blade is SPECIMEN greater than that in spheniscids with elongate blades. The condition in CADIC P 21 is clos- The new specimen, CADIC P 21, includes er to the condition in minor, for portions of the pelvis (primarily the left prea- example, which has a slightly shorter and cetabular ilium; ®g. 4a, b), nearly complete more expanded preacetabular ilium than do right femur (®g. 5), a fragment interpreted as other spheniscids. The anterior iliac blade in the head of the left femur (®g. 4a), nearly the fossil, however, appears more expanded complete right tibiotarsus (®gs. 6±8), and ®b- proximally than in Eudyptula. ula (®g. 9). The right femur lacks the pos- At least some expansion of the anterior terior portion of its proximolateral edge and preacetabular ilium is seen in all extant pen- posterior parts of the distal end (®g. 5). The guins other than forsteri and Ap- nearly complete right tibiotarsus is missing tenodytes patagonicus where the preaceta- the lateral cnemial crest and the medial half bular ilium is approximately equal in medio- of its distal end (®gs. 6±8). Overall, the spec- lateral width throughout its length. The an- imen is of an individual slightly smaller than terior portion of the ilium is also very little the largest extant penguin species, Aptenod- expanded in Eudyptes schlegeli, while in Eu- ytes forsteri, or the (table dyptes pachyrhynchus it is expanded proxi- 1). Compared to other fossil penguins, CAD- mally to approximately the same degree as IC P 21 is slightly smaller than the Eocene in all other spheniscid taxa considered except sp. (MoNZ 1449; Simpson, Eudyptula minor. The pelvic bones of fossil 2003 CLARKE ET AL.: FOSSIL PENGUIN 7

Fig. 4. The largest block of CADIC P 21 in (a, b) two views with parts of the acetabular and preacetabular ilium in ventromedial view, a small portion of the anterior pubis, the right tibiotarsus and the head of the left femur. Anatomical abbreviations: ac, acetabulum; f, femur, il, ilium; p, pubis; sc, scar for sacral transverse process; tb, tibiotarsus.

penguins are not well known; indeed, none than CADIC P 21, more like the condition appears so far to have been described in the in Aptenodytes forsteri, for example. literature. The anterior iliac blade appears to In the new fossil and in all spheniscids have only been pictured in one Paleogene studied other than individual specimens of penguin, a late Oligocene penguin from New antarctica and Pygoscelis papua, Zealand illustrated in Fordyce and Jones the ilium (and other pelvic bones) is unfused (1990: ®g. 18.6). In this illustration, the spec- to the sacral series. In the fossil penguin taxa imen appears to have a much narrower blade represented by the unpublished New Zealand 4 AMERICAN MUSEUM NOVITATES NO. 3423

Fig. 1. Simpli®ed geologic map of central and eastern Isla Grande, Tierra del Fuego, showing outcrops of Upper Cretaceous to Miocene synorogenic clastic wedges. The area of the penguin locality, at Cabo Campo del Medio±Punta Torcida, is bounded by a rectangle and detailed in the map provided in ®gure 2. late middle Eocene; and the Cerro Colorado The Punta Torcida Formation (ca. 220 m Formation, late middle to late Eocene. A ma- thick) is dominated by dysaerobic dark gray jor unconformity is recorded at the base of mudstones with sparse intercalations of light the Leticia Formation (Olivero and Malu- gray to yellowish ®ne turbidite sandstones. miaÂn, 1999; Olivero et al., 2002). At the northern limb of the anticline, it is covered by an unnamed sedimentary package (ca. 225 m thick) consisting of thick-bedded turbidite sandstones, tuffs, and mudstones. Collectively, these deposits are interpreted as parts of a turbidite system, including basinal thin-bedded turbidites in the basal part; base of slope thick-bedded turbidites in the middle part; and slope, distorted mudstones, and tuffs in the upper part (®gs. 2, 3). The ver- tical distribution of these facies association was related to the progradation of a fan-delta complex (Olivero and LoÂpez-Cabrera, 2001; Olivero et al., 2002). Planktonic foraminifera indicates a middle early Eocene age equiva- lent to the G. wilcoxensis Zone from New Fig. 2. Schematic map of the Cabo Campo del Zealand, or the tropical equivalent Zones P7 Medio±Punta Torcida anticline showing the distri- bution of the Punta Torcida Formation and unnamed to P8, or Zone AP6 from Antarctica (Olivero strata (lower Eocene) and the Leticia Formation (up- and MalumiaÂn, 1999). per middle Eocene). The star indicates where the Resting on a deeply incised unconformity, new specimen (CADIC P 21) was recovered. the vertebrate bearing Leticia Formation is 8 AMERICAN MUSEUM NOVITATES NO. 3423

Fig. 5. Right femur of CADIC P 21 prepared from the block in (left) anterior and (right) posterior views. Anatomical abbreviations: ail, anterior intermuscular line; cls, capital ligament scar; tcf, m. tibialis cranialis tendon fossa; ft, ®bular trochlea; mc, medial supracondylar crest; pf, popliteal fossa; tr, tro- chanteric surface. specimen (Fordyce and Jones, 1990: ®g. The capital ligament impression on the fem- 18.6) as well as in isolated sacral series from oral head is broad and deep. The proximal the Eocene of Seymour Island (Wiman, articular surface for the antitrochanter is 1905: ®g. 8, 1; Simpson 1971a), the sacral slightly concave and more U-shaped than V- series also appears to have been unfused (or shaped. By contrast, it is slightly more V- possibly incompletely fused) to the bones of shaped in other fossil penguins (e.g., Par- the pelvic girdle. aptenodytes antarcticus, AMNH 3338 in The right femur of CADIC P 21 was com- Simpson [1946]; Archaeospheniscus lowei, pletely prepared out of the matrix (®g. 5). OM C 47.27 in Simpson [1971b]) and com- 2003 CLARKE ET AL.: FOSSIL PENGUIN 9

Fig. 6. The CADIC P 21 block with the two articulating pieces of the right tibiotarsus joined. The tibiotarsus is shown in oblique anterolateral view. Portions of the left ilium, head of the left femur and part of the shaft of the right ®bula are also visible. Anatomical abbreviations: f, femur; fb, ®bula; il, ilium; tb, tibiotarsus. paratively strongly V-shaped in all extant The femoral shaft is nearly straight (®g. 5). penguins considered. In the V-shaped con- The posterior intermuscular line is weakly dition, the surface of the proximal end of the de®ned for the proximal half of the posterior femur in anterior or posterior view is ¯at to surface of the femur. It rises distally into a convex as it slopes gradually toward the di- strongly projected medial supracondylar minutive trochanteric crest from the base of crest that extends slightly less than one-third the humeral head. In CADIC P 21, this sur- the length of the femur. This crest crosses the face again curves up toward a diminutive tro- shaft from approximately its midpoint in pos- chanteric crest, but it creates a concave pro- terior view toward the medial edge of the ®le in anterior or posterior view. The CADIC medial condyle (®g. 5). No lateral supracon- P 21 condition was observed in procellari- dylar crest is developed. Of the distal femur iforms, and it is approached by the condition of CADIC P 21, posterior portions of the me- illustrated for Palaeeudyptes sp. (MoNZ dial condyle and anteroproximal portions of 1449; Simpson, 1971b). the lateral condyle are missing. The trochanteric crest appears to have The distal femur in CADIC P 21 is me- been relatively unprojected proximally or an- diolaterally narrow. Although parts of both teriorly. This morphology is indicated by in- condyles are missing, sections of the me- tact portions of the anterior tip of the proxi- diodistal and laterodistal surfaces of the fe- mal portion of the trochanteric crest and of mur are intact, and it is from the orientation the shaft just proximal and distal to the con- of these surfaces that a distal width only spicuously missing chip of the anterolateral slightly wider than the shaft can be inferred. surface (®g. 5). Proximal and posterior por- Furthermore, although the narrow ®bular tions of the crest are missing. However, a trochlea is proximolaterally incomplete, its deep muscular impression probably corre- distal and posterior edges are intact, from sponding to one of the obturator impressions which it is determined that it was little pro- (impressiones mm. trochanteris; Baumel and jected laterally. This inference is also consis- Witmer, 1993) is visible at what is inferred tent with the mediolaterally narrow, intact, to be approximately the crest's posterior end. proximal articular facet on the ®bula itself 10 AMERICAN MUSEUM NOVITATES NO. 3423

Fig. 7. The proximal portion of the CADIC P 21 tibiotarsus in oblique anterolateral view. The prominent ®bular crest and notch (foramen interosseum proximale; Baumel and Witmer, 1993) proximal to it are indicated. The anterior cnemial crest is unbroken, but the lateral is missing. Anatomical abbre- viations: acnc, anterior cnemial crest; fc, ®bular crest; la, lateral articular surface; n, notch.

(®g. 9). The distal end of the femur is an- the broken ends of these pieces of the tibi- teroposteriorly broad or uncompressed in dis- otarsus and at the chipped edge of the prox- tal view, which is the condition in other fos- imal femur is the internal structure of pelvic sil and extant pansphenisciforms. The pop- limb bones of CADIC P 21. These bones, liteal fossa is developed as a deep, discrete and the tibiotarsus in particular, are dense (or pit-shaped) depression located close to with an extremely thick layer of compact the lateral condyle. bone and very narrow medullary cavity. In- A poorly preserved fragment in the main deed, the tibiotarsal shaft in cross section ap- block of CADIC P 21 is interpreted as the pears nearly solid, a morphology developed head of the left femur in medial view (®g. in antarcticus and comment- 4a). It is the size of the head of the right ed on in Simpson (1946). While dense, less femur and bears a faint line in the same po- pneumatic, or apneumatic bones are devel- sition as the edge of the femoral head in the oped in a variety of crown clade mod- right. However, this fragment is heavily i®ed for diving (del Hoyo et al., 1992), the abraded, and no further morphology could be condition in CADIC P 21 is more extreme discerned including the presence or absence than even the condition in gaviids, for ex- of a capital ligament scar. ample. In fact, the condition would appear to The right tibiotarsus of CADIC P 21 is be matched in degree only by penguins and short; it is less than twice the length of the possibly other ¯ightless, highly modi®ed div- femur. This proportion is nearly the same as ing birds such as the extinct Great , Pin- that in Aptenodytes forsteri (table 1) and is guinus impennis, or the (Olson similar to that of other extant and fossil pen- and Hasegawa, 1996). Penguins have been guins (see Discussion). The tibiotarsus is cur- repeatedly remarked as distinguished by their rently preserved in two parts that articulate most extreme example of heavy or ``solid'' exactly (®g. 6). These pieces comprise one bones in crown clade birds (e.g., Simpson, still in the CADIC P 21 block (®g. 7) and 1946, Williams, 1995; del Hoyo et al., 1992). the second prepared out (®gs. 8). Visible at The shaft of the tibiotarsus is strongly 2003 CLARKE ET AL.: FOSSIL PENGUIN 11

Fig. 9. The ®bula of CADIC P 21 in lateral view. Anatomical abreviations: ff, facet for femur; sc, scar for attachment of lig. collateralis lateralis.

as well as in other ``Palaeeudyptines'' and other ``older fossil penguins'' (Simpson, 1957: 55), was stronger than that seen in any extant penguin taxa. Consistent with Simp- son's (1946, 1957) observations, the distal tibiotarsal shaft in CADIC P 21 is much more compressed than in any of the extant penguins compared. It is also, however, more compressed than in the late Oligocene or ear- ly Miocene Paraptenodytes antarcticus (AMNH 3338) from Patagonia. The morphology of the anterior cnemial crest is well preserved (®g. 7). It does not project extensively proximal to the articular surfaces for the femur. The patellar crest slightly overhangs the proximal surface of the tibiotarsus (®g. 7). On the anterior sur- face of the distal tibiotarsus, the extensor groove is relatively broad and slightly more medially located than in extant penguins (or Fig. 8. The distal tibiotarsus of CADIC P 21 in anterior view. Anatomical abreviation: ot, os- than in Paraptenodytes antarcticus; AMNH si®ed tendon. 3338). The ossi®ed supratendinal bridge is broken distally. Projecting proximally from under this bridge is a small element, which compressed anteroposteriorly. Marked com- may be part of an ossi®ed tendon (®g. 8). pression of the tibiotarsal shaft (distal to the The distal end of the tibiotarsus is de¯ected ventral end of the ®bular crest) has been not- medially (®g. 8). Additional morphologies of ed for other Eocene (Simpson, 1957) and late this region are not preserved. Oligocene±early Miocene (Simpson, 1946) The proximal portion of the right ®bula penguins. For example, Simpson (1957) was preserved in two pieces associated with commented that compression seen in a tibi- the tibiotarsus (®g. 6), one of which was pre- otarsus from the Eocene of Australia (Pa- pared off the CADIC P 21 block (®g. 9). It laeeudyptes cf. antarcticus; Simpson, 1957), is a robust element, and the distalmost pre- 12 AMERICAN MUSEUM NOVITATES NO. 3423

TABLE 1 Compared Measurements of CADIC P 21, Aptenodytes fosteri (AMNH 3745), and Paraptenodytes antarcticus (AMNH 3338)

served portion of the shaft shows no indi- IC P 21 is the fact that most fossil penguin cation of tapering markedly distally (®gs. 6, taxa have been described, named, and taxo- 9). In extant penguins the ®bula is also a ro- nomically revised based on morphologies of bust element that can extend most of the the humerus and tarsometatarsus (e.g., Simp- length of the tibiotarsus. The proximal artic- son, 1971b; Myrcha et al., 2002); these ele- ular facet for the femur is mediolaterally nar- ments are unpreserved in the new fossil. row (®g. 9). A raised scar for the lig. colla- Previous workers considered the best-pre- teralis lateralis is visible on the proximola- served parts of CADIC P 21, the femur and teral ®bular surface (®g. 9). tibiotarsus, to be comparatively undiagnostic elements in fossil and extant penguins (e.g., DISCUSSION Marples, 1952; Simpson, 1957; Grant-Mack- The scope of the current analysis was nec- ie and Simpson, 1973); thus, these elements essarily determined by the status of fossil have received relatively little commentary. penguin and of avian and, specif- Ilia (and other bones of the pelvic girdle), as ically, penguin systematics. The holotype well as the ®bulae of fossil penguins, are specimens of many previously named fossil even more poorly represented, and these el- penguin taxa are nonoverlapping single ele- ements of CADIC P 21 could be compared ments (or otherwise extremely fragmentary) to only one or two described or illustrated that cannot be compared with one another specimens. Thus, although CADIC P 21 was (e.g., Simpson, 1971a). Key fossil specimens compared to fossil and extant specimens (see (alternatively especially complete or de- appendix 1), and although its morphologies scribed as morphologically intermediate be- appear distinct from those known from these tween penguins and other avian taxa) have elements (see commentary in the Descrip- remained so far undescribed (see Fordyce et tion), we strongly believe that the decision to al., 1986; Fordyce and Jones, 1990). Addi- name the fossil as the holotype of a new tax- tionally limiting comparative study of CAD- on should be the result of a much-needed 2003 CLARKE ET AL.: FOSSIL PENGUIN 13 comprehensive, systematic revision of fossil dition, however, CADIC P 21 was differen- penguins, a project outside the scope of this tiated from all lineages of avian divers other paper. than penguins. A further issue determining the scope of The features considered at ®rst pass in the the current analysis is the paucity of avail- evaluation of the systematic position of able data on the phylogenetic position of the CADIC P 21 included the extreme bone den- penguin lineage in Aves, as well as data ad- sity, large size, and relative proportion of the dressing the relationships among extant pen- pelvic limb bones. The pelvic limb elements guins. The previously suggested sister taxa have a strikingly thick compact bone layer as of penguins are morphologically quite dis- developed only in birds modi®ed for pro- tinct. These potential outgroups of an anal- longed diving and only closely approached ysis of penguin relationships include a part by penguins of living taxa (see Description). of , or Procellariiformes as These heavy elements are also from an ani- a whole (FuÈrbringer, 1888; Simpson 1946, mal signi®cantly larger than any fossil or ex- 1971b; Sibley and Ahlquist, 1990), Gaviidae tant crown clade diving birds other than pen- (Olson, 1985), a Gaviidae ϩ Podicipedidae guins, the (Pinguinus impennis), clade (Cracraft, 1988), and possibly Phala- and the extinct plotopterids (Howard, 1969; crocoracidae (Paterson et al., 1995). Olson and Hasegawa, 1996). The specimen More recent phylogenetic results from is from an individual much larger than all analyses of mitochondrial gene sequence known fossil or extant taxa of , cor- data (Van Tuinen et al., 2001) and morpho- morants, including the large Mio- logical data (Livezey and Zusi, 2001) have cene , Meganhinga (Alvarenga, supported Procellariiformes and Gaviidae as 1995), or including the Eocene? ``Po- the successive outgroups of penguins. Groth larornis'' (Chatterjee, 1997; see Clarke and and Barrowclough (1999) also found support Chiappe, 2001), for example. for a clade including gaviids and spheniscids The femur is over one-half tibiotarsus (procellariiforms were not included) using length in CADIC P 21 (table 1; femur/tibi- nuclear gene sequence data. Nunn and Stan- otarsus: 0.63). To compare this ratio in pen- ley (1998) included multiple spheniscids as guins and other diving, and nondiving, birds, outgroups to analyses of procellariiform in- measurements given in appendix 2 of Gatesy terrelationships, and procellariiforms were and Middleton (1997) were used. The ratio not found to be paraphyletic with respect to for extant and fossil penguins ranged from penguins. Two morphology-based cladistic 0.55 to 0.70 (and most were between 0.60 analyses have been undertaken of extant pen- and 0.70). By contrast, in gaviids, the tibi- guins phylogenetic relationships (O'Hara, otarsus was conspicuously longer relative to 1989; Jones, 1995); however, neither has the femur (femur/tibiotarsus: 0.32±0.41; been published. The cladogram of O'Hara based on the ®rst ®ve measurements for Gav- (1989) was printed in Williams (1995) but ia immer and the ®rst four for Gavia stellata differs markedly from analyses of varying in Gatesy and Middleton [1997]), and in Pro- kinds of molecular data (Sibley and Ahlquist, cellariiformes the tibiotarsus was moderately 1990; Nunn and Stanley, 1998) and combin- longer (femur/tibiotarsus: 0.32±0.59; all from ing sequence data, isozyme, and behavioral Gatesy and Middleton [1997] were includ- characters (Paterson et al., 1995). ed), but the difference in length was less ex- Unfortunately, undertaking largescale treme. Other lineages of diving birds with analyses of the position of penguins within ratios approximating that of the penguins in- the avian crown clade and then optimizing clude phalacrocoracids, anhingids, sulids, al- apomorphies of penguins relative to their cids, and diving , while podicipedi- nearest outgroups were outside the scope of forms had a ratio closer to that of the gaviids the current project. Therefore, based on pre- (0.41±0.44). A ratio approximating that seen vious systematic results summarized above, in penguins is also seen in some basal avian Gaviidae and Procellariiformes were used as taxa such as , , and outgroups to hypothesize derived characters Tinamidae, as well as in a variety of other shared by CADIC P 21 and penguins. In ad- avian taxa. However, if Gaviidae and Pro- 14 AMERICAN MUSEUM NOVITATES NO. 3423 cellariiformes are the successive outgroups to gawa, 1996) approaches that seen in the fos- penguins (e.g., Van Tuinen et al., 2001; Liv- sil. Whether a more expanded anterior iliac ezey and Zusi, 2001), a higher ratio could be blade is ancestral to pansphenisciforms de- optimized as a local of pen- serves further scrutiny. The anteroposteriorly guins, including the fossil. compressed tibiotarsal shaft seen in CADIC Further morphologies with restricted dis- P 21 is most closely matched by phalacro- tributions among the surveyed taxa consis- coracids and anhingids of extant taxa, al- tent with identi®cation of the fossil as a pen- though this condition is also developed in guin include several characters of the pelvic fossil penguins (e.g., Simpson, 1946), as girdle and limb. For example, incomplete or mentioned in the description. Slight com- complete lack of coossi®cation of sacrum pression is seen in some procellariiforms. and pelvic bones occurs in CADIC P 21, Again, whether the state in CADIC P21 and penguins, some procellariiforms, gaviids, other fossil penguins is ancestral to Pan- podicipedids, alcids, and a variety of other sphenisciformes merits additional study. avian taxa but not usually in phalacrocora- In Paraptenodytes antarcticus (AMNH cids, anhingids, sulids, or apparently in plo- 3338) and CADIC P 21, the medial supra- topterids (Olson and Hasegawa, 1996). Fur- condylar crest is well projected but extends thermore, in CADIC P 21 and penguins, the proximally markedly less than one-half the distal femur is narrow mediolaterally and the length of the femur; this condition was also ®bular trochlea is relatively unprojected approached by Eudyptula minor of Sphenis- (with a narrow corresponding proximal facet cidae. By contrast, in Pygoscelis adeliae, Py- on the ®bula); by contrast, in podicipedids, goscelis antarctica, Aptenodytes forsteri, and gaviids, anhingids, and phalacrocoracids, the Aptenodytes patagonicus the crest extends distal femur is broader with a projected ®b- proximally one-half the length of the femur. ular trochlea. (In procellariiforms and sulids, The Spheniscus, Eudyptes, and the distal femur was only slightly broader species surveyed had a more distally restrict- than the penguin condition.) CADIC P 21 ed crest than did the Pygoscelis and Apten- also shares with penguins relative to other odytes species but one more proximally ex- diving taxa, including gaviids and procellar- tensive than CADIC P 21, Paraptenodytes iiforms, a comparatively low anterior cne- antarcticus (AMNH 3338), and Eudyptula mial crest on the tibiotarsus. minor. The distribution of this morphology The smoothly concave pro®le of the prox- imal surface of the femur in either lateral or suggests that a more distally restricted medial medial view in CADIC P 21 is not developed supracondylar crest could be plesiomorphic in any extant penguins but appears ap- for Spheniscidae within Pansphenisciformes; proached by some fossil penguins (see dis- however, further comparisons are necessary, cussion in the Description), and is also seen particularly to assess variation in this char- in Procellariiformes (but not in gaviids, pod- acter for other fossil outgroups of Sphenis- icipedids, anhingids, or phalacrocoracids). cidae. The extensor groove on the tibiotarsus, more medially located in CADIC P 21 than in ex- CONCLUSIONS tant, and other fossil, penguins, is also more medially located in procellariiforms, sug- The identi®cation of parts of Pansphenis- gesting that this condition could be found to ciformes, or the penguin stem clade, in the be plesiomorphically retained in basal Pan- middle (e.g., CADIC P 21) and even early sphenisciformes. (Fordyce et al., 1986) Eocene constrains the The anterior expansion of the preacetabu- timing of the divergence of the penguin lin- lar iliac blade seen in CADIC P 21 is better eage from its nearest sister taxon. A long developed than in extant penguins but is ap- overdue phylogenetic analysis including fos- proached by Eudyptula minor (see Descrip- sil and extant penguins may place Paleogene tion). Expansion is slight in procellariiforms, taxa within the penguin crown clade. How- while the condition in phalacrocoracids, an- ever, without evidence of such af®nity these hingids, and plotopterids (Olson and Hase- do not address the timing of any of 2003 CLARKE ET AL.: FOSSIL PENGUIN 15 the divergences within Spheniscidae or the Mayr and Ewan Fordyce improved the man- origin of the penguin crown as a whole. uscript. M.I. LoÂpez-Cabrera (CADIC) helped Indeed, to give further pause to con¯ating with the ®eldwork at the new locality. Sup- the age of earliest fossil penguin with a min- port for part of this research by PIP 4304- imum age of the origin of Spheniscidae, CONICET and PICT 07-8675-ANPCYT- there is evidence supporting placement of BID 1201/OC-AR, Argentina, to E.B. Oliv- CADIC P 21 and other Paleogene taxa out- ero, and by Section of Vertebrate Paleontol- side the penguin crown. ogy Frick Fund of the American Museum of This result is consistent with Simpson's Natural History to J.A. Clarke is gratefully (1946, 1971b) conclusion that pre-Miocene acknowledged. (and some Miocene) fossil taxa were not closely related to any extant penguins; these REFERENCES taxa were placed outside Simpson's Sphen- iscidae (consistent with the usage of that Acosta Hospitaleche, C., C.P. Tambussi, and M.A. name here; Simpson, 1946, 1971b). Mor- Reguero. 2001. CataÂlogo de tipos de aves foÂs- phologies of the quadrate, humerus, and tar- iles del Museo de La Plata. Revista del Museo sometatarsus present in Eocene and Miocene de La Plata Serie TeÂcnica y DidaÂctica, 41: penguins and absent in extant penguins have 1±28. been given to support (Simpson, 1946: 80± Alvarenga, H.M.F. 1995. A large and probably 82, 1971b: 367) this conclusion. In addition, ¯ightless Anhinga from the Miocene of Chile. the marked anteroposterior compression of Courier Forschungsinstitut Senckenberg 181: 149±161. the tibiotarsus in Eocene taxa (including Baumel, J.J., A.S. King, J.E. Breazile, H.E. CADIC P 21) and some Miocene taxa Evans, and J.C. Vanden Berge (editors). 1993. (Simpson, 1946, 1957) but not in any extant Handbook of avian anatomy: Nomina Anatom- taxon is consistent with the placement of ica Avium, 2nd ed. Cambridge, MA: Nuttall Or- these fossil penguins outside Spheniscidae. nithological Club, 779 pp. Further understanding of the timing of Baumel, J.J., and L.M. Witmer. 1993. Osteologia. pansphenisciform divergences will come In J.J. Baumel, A.S. King, J.E. Breazile, H.E. from combined phylogenetic analyses of fos- Evans, and J.C. Vanden Berge (editors), Hand- sil and extant penguin taxa. CADIC P 21 book of avian anatomy: Nomina Anatomica Av- clearly indicates, however, the presence of ium, 2nd ed.: 23: 45±132.Cambridge, MA: Nut- penguins in southernmost South America ap- tall Ornithological Club. proximately 20 million years earlier than pre- Cantino, P.D., and K. de Queiroz. 2000. Phylo- viously supported. In this regard, the speci- Code: a phylogenetic code of biological no- menclature (online paper). Athens: Ohio Uni- men offers the important potential for insight versity (updated 3 May 2000). Available at into penguin paleobiogeography with future www.ohiou.edu/phylocode/. work. Case, J.A. 1992. Evidence from fossil for a rich Eocene Antarctic marine environ- ACKNOWLEDGMENTS ment. The Antarctic paleoenvironment: a per- spective on global change. Antarctic Research We thank Rick Edwards for providing ex- Series 56: 119±130. quisite digital photographs of the specimen, Chatterjee, S. 1997. The rise of birds. Baltimore: Paul Sweet and Ben Burger for assistance Johns Hopkins Press, 312 pp. with collections of the Sections of Ornithol- Cione, A.C., R.A. del Valle, C.A. Rinaldi, and E.P. ogy and Vertebrate Paleontology (AMNH), Tonni, 1976. Nota preliminar sobre los pingui- and L. Gliubich (Estancia La Fueguina) for nos y tiburones del Terciario inferior de la Isla Vicecomodoro Marambio, AntaÂrtida. Contri- assistance and logistical support in the ®eld. buciones del Instituto AntaÂrtico Argentino 213: We also thank Marilyn Fox for assistance 3±21. with the preparation of the new specimen and Clark, G.A., Jr. 1993. Termini situm et directi- the Yale Peabody Museum for hosting P. onem partium corporis indicantes. In J.J. Bau- Puerta during its preparation. Ewan Fordyce mel, A.S. King, J.E. Breazile, H.E. Evans, and provided valuable information about New J.C. Vanden Berge (editors), Handbook of avi- Zealand penguins. Comments by Gerald an anatomy: Nomina Anatomica Avium, 2nd ed. 16 AMERICAN MUSEUM NOVITATES NO. 3423

23: 1±5. Cambridge, MA: Nuttall Ornithologi- Glaessner, M.F. 1955. Pelagic fossils (Aturia, pen- cal Club. guins, whales) from the Tertiary of South Aus- Clarke, J.A., and L.M. Chiappe. 2001. A new car- tralia. Records of the South Australian Museum inate bird from the Late Cretaceous of Pata- 11: 353±372. gonia (Argentina). American Museum Novita- Grant-Mackie, J.A., and G.G. Simpson. 1973. tes 3323: 1±23. Tertiary penguins from the North Island of New Clarke, J.A., and M.A. Norell. 2002. The mor- Zealand. Journal of the Royal Society of New phology and phylogenetic position of Apsaravis Zealand 3: 441±452. ukhaana from the Late Cretaceous of Mongo- Groth, J.G., and G.F. Barrowclough. 1999. Basal lia. American Museum Novitates 3387: 1±46. divergences in birds and the phylogenetic util- Cracraft, J.L. 1988. The major of birds. In ity of the nuclear RAG-1 gene. Molecular Phy- M.J. Benton (editor), The phylogeny and clas- logenetics and Evolution 12: 115±123. si®cation of tetrapods 1(35A): 339±361.New Howard, H. 1969. A new avian from Kern Coun- York: Clarendon Press. ty, California. Condor 71: 68±69. de Queiroz, K., and J. Gauthier. 1990. Phylogeny Jenkins, R.J.F. 1974. A new giant penguin from as a central principle in taxonomy: phylogenet- the Eocene of Australia. Palaeontology 17: ic de®nitions of taxon names. Systematic Zo- 291±310. ology 39: 307±322. Jones, C.M. 1995. Systematics of a new fossil de Queiroz, K., and J. Gauthier. 1992. Phyloge- penguin (Spheniscidae) from the Kokoamu netic taxonomy. Annual Review of Ecology Greensand (Duntroonian Stage, Upper Oligo- and Systematics 23: 449±480. cene), South Island, New Zealand. Unpublished del Hoyo, J., A. Elliott, and J. Sargatal (editors). MSc thesis, University of Otago. 1992. Handbook of the birds of the world, vol. Kennedy, M., and R.D.M. Page. 2002. su- 1, Barcelona: Lynx Edicions. pertrees: combining partial estimates of procel- Dingle, R.V., and M. Lavelle. 1998. Late Creta- lariiform phylogeny. Auk 119(1): 88±108. ceous±Cenozoic climatic variations of the Livezey, B.C., and R.L. Zusi. 2001. Higher- northern : new geochemical phylogenetics of modern Aves based on com- evidence and review. Palaeogeography, Palaeo- parative anatomy. Netherlands Journal of Zo- climatology, Palaeoecology 141: 215±232. ology 51(2): 179±205. Fordyce, R.E. 1991. A new look at the fossil ver- MalumiaÂn, N., A. Ardolino, M. Franchi, M. Re- tebrate record of New Zealand. In P. Vickers- mesal, and F. Salani. 1999. La sedimentacioÂn y Rich et al. (editors), Vertebrate palaeontology el volcanismo terciarios en la Patagonia Ex- of Australasia: 1191±1316. Melbourne, Austra- traandina. In R. Caminos (editor), GeologõÂa Ar- lia: Pioneer Design Studio and Monash Uni- gentina Instituto de GeologõÂa y Recursos Mi- versity. nerales, Anales 29(18): 557±612. Fordyce, R.E., and C.M Jones. 1990. The history MalumiaÂn, N., E. B. Olivero, and A. Concheyro. of penguins and new fossil penguin material 1994. Eocene microfossils from the Leticia For- from New Zealand. In L.S. Davis and J.D. Dar- mation, Tierra del Fuego Island, Argentina. by (editors), Penguin biology: 419±446. San Ameghiniana 31: 398. Diego: Academic Press. Marenssi, S.A., S.N. Santillana, and C.A. Rinaldi. Fordyce, R.E., C.M. Jones, and B.D. Field. 1986. 1998. Stratigraphy of the La Meseta Formation The world's oldest penguin? Geological Society (Eocene), Marambio (Seymour) Island, Antarc- of New Zealand Newsletter 74: 56±57. tica. AsociacioÂn PaleontoloÂgica Argentina Pub- FuÈrbringer, M. 1888. Untersuchunge zur Mor- licacioÂn Especial 5, Paleogeno de AmeÂrica del phologie und Systematik der VoÈgel. Amster- Sur y de la PenõÂnsula AntaÂrtica: 137±146. dam: Holkema, 1751 pp. Marples, B.J. 1952. Early Tertiary penguins of Gatesy, S.M., and K.M. Middleton. 1997. Biped- New Zealand. New Zealand Geological Survey alism, ¯ight and the evolution of theropod lo- Paleontological Bulletin 20: 1±66. comotor diversity. Journal of Vertebrate Pale- Myrcha, A., P. Jadwiszczack, C.P. Tambussi, J.I. ontology 17(2): 308±329. Noriega, A. Tatur, A. Gadzdzicki, and A. del Gauthier, J., and K. de Queiroz. 2001. Feathered Valle. 2002. Taxonomic revision of Eocene , ¯ying dinosaurs, crown dinosaurs penguins based on tarsometatarsal morphology. and the name ``Aves''. In J. Gauthier and L.F. Polish Polar Research 23(1): 5±46. Gall (editors), New perspectives on the origin Nunn, G.B., and S.E. Stanley. 1998. Body size and early . Proceedings of the effects and rates of cytochrome b evolution in International Symposium in Honor of John H. tube-nosed . Molecular Biology and Ostrom: 7±41. New Haven, CT: Peabody Mu- Evolution 15(10): 1360±1371. seum of Natural History. O'Hara, R.J. 1989. An estimate of the phylogeny 2003 CLARKE ET AL.: FOSSIL PENGUIN 17

of the living penguins (Aves: Spheniscidae). ships. Palaeogeography, Palaeoclimatology, Pa- American Zoologist 29: 11A. [abstract; clado- laeoecology 2776: 1±22. gram reprinted in T. Williams, 1995] Sadler, P.M. 1988. Geometry and strati®cation of Olivero, E.B., and C.M.I. LoÂpez-Cabrera. 2001. the uppermost Cretaceous and Paleogene units Ichnology of syntectonic turbidite, estuarine of Seymour Island, northern Antarctic Penin- and shelf deposits. Eocene, Punta Torcida an- sula. In Geology and paleontology of Seymour ticline, Isla Grande de Tierra del Fuego. IV Re- Island,Antarctic Peninsula: 169: 303±320. Geo- unioÂn Argentina de IcnologõÂa y II ReunioÂn de logical Society of America Memoir. IcnologõÂa del Mercosur, Septiembre 2001, Tu- Sibley, C.G., and J.E. Ahlquist. 1990. Phylogeny cumaÂn, Argentina. ResuÂmenes: 62. and classi®cation of birds: a study in molecular Olivero, E.B., and N. MalumiaÂn. 1999. Eocene evolution. New Haven: Yale University Press, stratigraphy of southeastern Tierra del Fuego, 976 pp. Argentina. American Association of Petroleum Simpson, G.G. 1946. Fossil penguins. Bulletin of Geologists Bulletin 83(2): 295±313. the American Museum of Natural History 87: Olivero, E.B., and N. MalumiaÂn. 2002. Upper 1±99. Cretaceous±Cenozoic clastic wedges from the Simpson, G.G. 1957. Australian fossil penguins, Austral±Malvinas foreland basins, Tierra del with remarks on penguin evolution, and distri- Fuego, Argentina: eustatic and tectonic con- bution. Records of the South Australian Mu- trols. 3rd European Meeting on the Palaeontol- seum 13: 51±170. ogy and Stratigraphy of Latin America (EMPS- Simpson, G.G. 1971a. Review of the fossil pen- LA), 19±20 September, 2002. Toulouse, guins from Seymour Island. Proceedings of the : University Paul Sabatier. Royal Society (London), ser. B 178: 357±387. Olivero, E.B., N. MalumiaÂn, and G. Jannou. 2002. Simpson, G.G. 1971b. A review of the pre-Plio- EstratigrafõÂa y facies de sistemas turbidõÂticos, cene penguins of New Zealand. Bulletin of the estuaÂricos y de plataforma, Eoceno, Andes Fue- American Museum of Natural History 144: guinos, Argentina. IX ReunioÂn Argentina de 321±378. SedimentologõÂa, 2123 May, Cordoba, Argenti- Simpson, G.G. 1972. Conspectus of Patagonian na, ResuÂmenes: 57. fossil penguins. American Museum Novitates Olson, S.L. 1985. The fossil record of birds. In D.S. Farner, J.R. King, and K.C. Parkes (edi- 2488: 1±37. tors), Avian biology: 8: 79±238. New York: Simpson, G.G. 1981. Notes on some fossil pen- Academic Press. guins, including a new from Patagonia. Olson, S.L., and Y. Hasegawa. 1996. A new genus Ameghiniana 18(3±4): 266±272. and two new species of gigantic Plotopteridae Van Tuinen, M. Butvill, D.B., J.A.W. Kirsch, and from Japan (Aves; ). Journal of S.B. Hedges. 2001. Convergence and diver- Vertebrate Paleontology 16(4): 742±751. gence in the evolution of aquatic birds. Pro- Paterson, A.M., G.P. Wallis, and R.D. Gray. 1995. ceedings of the Royal Society (London), ser. B Penguins, , and parsimony: does cladistic 268: 1345±1350. analysis of behavior re¯ect seabird phylogeny. Williams, T. 1995. The penguins. Oxford: Oxford Evolution 49(5): 974±989. University Press, 295 pp. Reguero, M.A., S.A. Marenssi, and S.N. Santil- Wiman, C. 1905. UÈ ber die alttertiaÈren Vertebraten lana. 2002. Antarctic Peninsula and South der Seymour-Insel. Wissenschaftliche Ergebnis- America (Patagonia) Paleogene terrestrial fau- se der Schwedischen SuÈdpolar Expedition nas and environments: biogeographic relation- 1901±1903 3(1): 1±37.

APPENDIX 1 Specimens compared to the new fossil in the Eudyptula minor, AMNH 6257. Procellariiformes: preparation of the description are as follows: (Diomedeidae) Phoebastria (Diomedea) irrorata, Sphenisciformes: Pygoscelis adeliae, AMNH AMNH 1628; Thalassarche (Diomedea) cauta, 3649; Pygoscelis antarctica, AMNH 26160; Py- AMNH 1436; (Procellariidae) Macronectes gigan- goscelis papua, AMNH 22679; Spheniscus demer- teus, AMNH 5400; (Hydrobatidae) Fregetta tro- sus, AMNH 12782; Spheniscus humboldti, AMNH pica, AMNH 5330; Oceanodroma leucorhoa, 4920; Aptenodytes forsteri, AMNH 3745; Apten- AMNH 22019; (Pelecanoididae) Pelecanoides gar- odytes patagonicus, AMNH 1624; Megadyptes an- notii, AMNH 3125; Pelecanoides magellani, tipodes, AMNH 5613; Eudyptes schlegeli, AMNH AMNH 23569; : Gavia immer, 5399; Eudyptes pachyrhynchus, AMNH 26509; AMNH 26310; Gavia stellata AMNH 6971. Pod- 18 AMERICAN MUSEUM NOVITATES NO. 3423 icipediformes: Podiceps grisegena, AMNH 3878. The new specimen was also compared with fos- ``Pelecaniformes'': (Anhingidae) Anhinga rufa, sil penguin material illustrated in the literature AMNH 5240; Anhinga anhinga, AMNH 2919; (e.g., Simpson, 1946, 1957, 1971a, 1971b; Wi- (Phalacrocoracidae): Stictocarbo (Phalacrocorax) man, 1905), plotopterid (Plotopteridae: ``Pelecan- magellanicus, AMNH 23560; Hypoleucos (Phala- iformes''; Howard, 1969) material illustrated in crocorax) varius, AMNH 5412; (Sulidae): Morus Olson and Hasegawa (1996), and Meganhinga (Sula) bassanus AMNH 3127. chilensis as described by Alvarenga (1995). Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates and Bulletin pub- lished during the last ®ve years are avail- able at World Wide Web site http://li- brary.amnh.org. Or address mail orders to: American Museum of Natural History Li- brary, Central Park West at 79th St., New York, NY 10024. TEL: (212) 769-5545. FAX: (212) 769-5009. E-MAIL: sci- [email protected]

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