Associated penguin remains found in Bartonian levels of the Submeseta Formation (Seymour Island, Antarctica), including cranium and mandible, both partial tarsometatarsi, and some other fragmentary bones, are analyzed here. This specimen preserves the first cranium reliably assigned to the giant form Anthropornisgrandis, and constitutes the first opportunity to taxonomically assign a cranial material to any of the Antarctic penguin species. A discussion of the diet preferences and feeding mechanisms of A. grandis is supported here by three-dimensional paleoneurological and cranial-jaw muscular reconstructions. We propose that A. grandis was a penguin with a voluminous musculature strongly attached to the neck and skull, adapted to chase and hunt fish during diving.
Les restes associés aux manchots trouvés dans les niveaux de Bartonien dans la formation de Submeseta (île Seymour, Antarctique), y compris le crâne et la mandibule, des tarsométatarses partiels et quelques autres os fragmentaires, sont analysés ici. Ce spécimen conserve le premier crâne attribué de manière fiable à la forme géante Anthropornisgrandis et constitue la première occasion d’affecter de manière taxonomique un matériau crânien à l’une des espèces de manchots antarctiques. Une discussion sur les préférences alimentaires et les mécanismes d’alimentation d'A. grandis est étayée ici par des reconstructions paléoneurologiques en trois dimensions et une reconstruction musculaire crâne–mâchoire. Nous proposons qu’A. grandis était un manchot à la musculature volumineuse, fortement attachée au cou et au crâne, adapté à la poursuite et à la chasse des poissons en plongée.
Fossil penguin, Anatomy, Paleoneurology, Paleobiology, Submeseta Formation, Eocene, AntarcticaManchot fossile, Anatomie, Paléoneurologie, Paléobiologie, Formation de Submeseta, Éocène, AntarctiquepresentedHandled by Michel LaurinIntroduction
Anthropornis Wiman, 1905 is a genus described more than a century ago and one of the most remarkable Eocene penguins found in Antarctica. The robustness and massiveness of its postcranial bones caught the attention of Carl Wiman in 1905, and made it worthy of the name Anthropornis, meaning “man-bird” (Wiman, 1905a). For many years, Anthropornis was considered the largest penguin ever, until giant isolated tarsometatarsi of Palaeeudyptes klekowskii were discovered in the Eocene of Antarctica (Acosta Hospitaleche, 2014). Regardless of size, Anthropornis remained an intriguing taxon because, except for the tarsometatarsus, other skeletal elements were barely known (Jadwiszczak, 2006, Jadwiszczak, 2012, Marples, 1953 and Wiman, 1905a).
Initially, Wiman, 1905a and Wiman, 1905b described the holotype tarsometatarsi of Anthropornis and Pachypteryx (junior synonym of the former; Myrcha et al., 2002 and references therein), focusing on the striking size of the material, and the relative size and location of the foramina vascularia. This, however, did not include a differential diagnosis of each genus. Later on, some other non-homologous and isolated elements included in the size groups 1 to 4 proposed by Wiman (1905b) were assigned to Anthropornis nordenksjoeldi (Marples, 1953), probably based on their size and robustness. Although now obsolete, Wiman's size groups allowed the classification of penguin remains using non-conventional categories (Wiman, 1905b). More recently, an important systematic revision based on the tarsometatarsus provided a revised diagnosis in which Anthropornis was unambiguously differentiated from any other penguin (Myrcha et al., 2002).
Currently, and even though there were at least two species recognized in an attempt to analyze its taxonomic diversity (Acosta Hospitaleche et al., 2013 and Jadwiszczak, 2006), the number of species of Anthropornis is still under debate (Acosta Hospitaleche et al., 2017 and Jadwiszczak, 2006). The possibility that a third species has lived sympatrically with A. grandis and A. nordensjkoeldi in Antarctica during the Eocene is not discarded (see Jadwiszczak, 2006). Only isolated appendicular bones were described for both species of Anthropornis, until the finding of a partially complete skeleton assigned to Palaeeudyptes gunnari (Acosta Hospitaleche and Reguero, 2010) allowed the reinterpretation of limb elements of Anthropornis sp. (Jadwiszczak, 2012). Although Marples (1953) stated that these remains may not belong to the same skeleton, a re-examination of the limbs allowed Jadwiszczak (2012) to confidently assign them to a single individual of Anthropornis sp.
The Eocene Marambio/Seymour Island (Fig. 1) penguin record includes isolated cranial elements (Table 1) with a systematic allocation restricted to the family level. On the other hand, the few incomplete skeletons described from the Eocene of Antarctica (Acosta Hospitaleche, 2016, Acosta Hospitaleche and Di Carlo, 2010, Acosta Hospitaleche and Reguero, 2010, Acosta Hospitaleche and Reguero, 2014 and Jadwiszczak, 2012) do not preserve the skull.
Here, new associated remains of a single individual (Fig. 2) with cranium, mandible (Fig. 3 and Fig. 4), both partial tarsometatarsi (Fig. 5), and some other fragmentary bones are described. The remains were found within Bartonian levels of the Submeseta Formation (Seymour Island), Antarctica (Fig. 1). This partial skeleton has the first cranium reliably assigned to Anthropornis, or to any Antarctic penguin species ever. A detailed description of MLP 14-XI-27-84, including osteology, paleoneurology, myological reconstructions, and comments on the inferred paleobiology is provided.
MLP 14-XI-27-84 was collected by members of the Museo de La Plata (Universidad Nacional de La Plata) and of the Instituto Antártico Argentino during the summer fieldtrip (2013–2014) to Seymour Island, Antarctica. Bones were mechanically prepared and housed at the División Paleontología Vertebrados of the Museo de La Plata.
Terminology
Osteological and myological descriptions follow the anatomical terminology proposed by Baumel et al. (1993), complemented by Livezey and Zusi (2006) when necessary. Since MLP 14-XI-27-84 constitutes the first skull assigned to Anthropornis grandis with certainty, comparisons with other Paleogene penguins are also given. Measurements were taken with a Vernier caliper of 0.1 mm of increment.
Anatomical reconstructions
A three-dimensional external scan was performed with a NextEngine 3D Laser Scanner Ultra HD from the División Paleontología Vertebrados (Museo de La Plata). A retro-deformation of the mesh was performed in Landmark Editor (Wiley et al., 2007) in order to partially revert the strong taphonomic deformation of the skull (Fig. 6 and Fig. 7). A microCT scan of the cranium was performed at the Y-TEC facilities (La Plata, Argentina) using a Bruker SkyScan 1173 micro-tomographer, and applying a voltage of 120 kV and a current of 66 μA, resulting in 1117 slices with a thickness of 70 μm. The segmentation of the brain and inner ear was made using the software Materialise Mimics (18.0).
Geographic and stratigraphic context
The fossil site IAA 4/12, from where MLP 14-XI-27-84 was collected, is located in the northern sector of Seymour/Marambio Island (Fig. 1a). This island is in the Weddell Sea near the end of the Antarctic Peninsula (Fig. 1c), and remains free of ice during the summer, which makes it one of the best places for paleontological fieldwork in Antarctica. Sediments on Seymour Island are Late Cretaceous to late Eocene/?earliest Oligocene in age, and belong to the lower Marambio Group (composed of the López de Bertodano and Sobral formations) and the overlying Seymour Island Group (Cross Valley, La Meseta, and Submeseta formations) filling the James Ross Basin (Montes et al., 2013).
The specimen MLP 14-XI-27-84 was found within a single block cropping out in surficial deposits of the Submeseta II Allomember of the Submeseta Formation (Montes et al., 2013). These strata are also known as level 38 (Montes et al., 2013), or Telm 7 (Sadler, 1988), and represent the richest unit in terms of fossil penguin diversity and abundance (Fig. 1b, d).
Systematic paleontology
Order Sphenisciformes Sharpe, 1891
Anthropornis Wiman, 1905
Anthropornis grandis (Wiman, 1905)
Fig. 3, Fig. 4 and Fig. 5
Material. MLP 14-XI-27-84, partial skeleton including the cranium, mandible, tarsometatarsi, pedal phalanges, several pieces of vertebrae, ribs, and pygostyle.
Locality and horizon. IAA 4/12 locality (Seymour Island), Antarctic Peninsula, West Antarctica (Fig. 1a, c). Priabonian levels of the Submeseta Formation, James Ross Basin (Fig. 1b, d).
Description and comparisons
Bones are robust and show signs of the osteosclerosis typical of penguins (Houssaye, 2009, Ksepka et al., 2012a and Meister, 1962). The cranium is partially deformed (Fig. 3 and Fig. 5); it includes the complete calvaria and the basis cranii externa, part of the ossa cranii and the cavum tympanicum, missing the rostrum and palate. The mandible is almost complete.
Both right and left tarsometatarsi are preserved. The left one is complete (Fig. 5), whereas the right one is only represented by a piece of the second and third metatarsal. Other associated elements include fragments of ribs (some of them with fused uncinate processes), cervical and thoracic vertebrae, postzigapophysis of cervical and thoracic vertebrae, pygostyle, six proximal phalanges plus one phalanx ungualis (Fig. 5c), and several undetermined pieces of bones.
Cranium:Anthropornis grandis MLP 14-XI-27-84 is similar in size to MLP 12-I-20-2 (see Table 1 for the complete list of Antarctic crania), but larger than Muriwaimanu (Paleocene of New Zealand, Mayr et al., 2017 and Slack et al., 2006), Perudyptes (Ksepka et al., 2010), Icadyptes (Ksepka et al., 2008, although high deformation of the latter makes comparison more difficult), Inkayacu (Eocene of Peru, Clarke et al., 2010), or any of the crania found before in Antarctica. The roof of the cranium and the interorbital region are wider and more robust than those in MLP 12-I-20-2, which is the most similar Antarctic cranium (Acosta Hospitaleche, 2013: fig. 5).
The skull roof is a little crushed and the interorbital region is not perfectly aligned with the crista nuchalis sagittalis due to dorso-lateral deformation (Fig. 3a, c). The retrodeformed three-dimensional mesh — obtained by the selection of symmetrical points (Fig. 6) — shows an approximation of the skull prior to deformation (Fig. 7). The bridge between both orbitae keeps the same width up to the os lacrimale. Neither the suture between the processus frontalis nasalis and the processus frontalis premaxilaris nor the suture between right and left processes are visible (Fig. 3a).
Both fossae glandulae nasale are shallow furrows with a rugose texture that narrow cranially (indeed, the fossae glandulae nasale are relatively narrow in all the Antarctic crania). These fossae are medially demarcated by a sharp ridge, which is lower at the caudal end. There is no supraorbital edge at the lateral margin (Fig. 3a). A similar configuration is observed in Icadyptes, Perudyptes, Inkayacu, Muriwaimanu, and Sequiwaimanu.
The orbits are cranio-caudally expanded and limited by a slender and ventro-laterally projected processus postorbitalis (which is wider cranio-caudally in Sequiwaimanu) and an enlarged processus supraorbitale lacrimale (Fig. 3a, c, f). The latter process has a rounded tip that diverges from the sagittal plane toward the caudal end and over the orbit as an expansion of the lacrimal foot (sensuCracraft, 1968). Ventrally, the processus postorbitalis is divided by two fossae.
The anterior fossa is postero-medially more expanded and deeper than in MLP 12-I-20-1. Ventrally to this fossa, the scar for the m. pseudotemporalis superficialis, in the area aspera of the os laterosphenoidale, is ventrally more extended in A. grandis than in other Eocene forms. Beneath this scar, there is another one that includes a fossa deeper than the one of MLP 12-I-20-1. The posterior fossa is also deep, triangular, but rather short in comparison with the former (Fig. 3f). The total area of insertion of the m. pseudotemporalis superficialis seems to have been deep, large, and postero-medially more extended than in MLP 12-I-20-1.
The head of the lacrimal is cranio-caudally elongated and, approximately at the middle point of the lacrimal head, the descending process extends ventrally, but also caudally. It ends in an enlarged plate (foot sensuCracraft, 1968) expanded in a processus supraorbitale lacrimale (see Cracraft, 1968 for subdivisions of the lacrimal) (Fig. 3a, f). This condition is strongly similar to that observed in Procellariiformes (CAH, pers. obs.).
Although the lacrimal is well developed in all Spheniscidae, the striking enlargement of the processus supraorbitale lacrimale appears only in Anthropornis grandis, Icadyptessalasi (Ksepka et al., 2008: fig. 5A), and Waimanu tuatahi (Slack et al., 2006: fig. 1Ab). This projection is not developed in Kairuku (Oligocene of New Zealand, Ksepka et al., 2012a), and it cannot be evaluated in Inkayacu, or Sequiwaimanu. In most penguin species (e.g., Spheniscus) the descending process of the lacrimal is notably shorter than in A. grandis. Otherwise, the development of the lacrimal plate varies among the species; in Aptenodytes forsteri for example, it is larger than in other living genera, but shorter than in A. grandis.
The naso-frontal flexure, absent in Icadyptes (according to the description of Ksepka et al., 2008) and Inkayacu, is present in A. grandis. The depression that typically appears in this zone, at least in some extant penguins, is present in A. grandis, Sequiwaimanu, and seems also present in Perudyptes.
The crista nuchalis sagittalis is straight and sharp. It separates both fossae temporale that extend dorsally reaching the sagittal midline. The whole complexes of cristae nuchale draw an inverted U-shaped fossa temporalis in dorsal view. At its dorsal end, the crista nuchalis transversa constitutes a higher wall compared with the fossa nuchalis temporalis (Fig. 3a).
The fossa temporalis is slightly crushed, mainly on the left side, but doubtlessly constitutes a wide and deep place for muscular attachment for the m. adductor mandibulae externus (Fig. 3a). The fossa temporalis is similarly extended in all the materials and is limited by strong borders as in MLP 12-I-20-1, MLP 12-1-20-2, Icadyptes, Perudyptes, Inkayacu, and Muriwaimanu. Contrarily, in Sequiwaimanu the fossa temporalis seems restricted to a more caudal position. In all the other crania compared here, the configuration is similar: the fossae meet each other dorsally at the sagittal line and the strong cristae nuchale limits them cranially and caudally. The os supraoccipitale extends dorsally, caudal to the crista nuchalis sagittalis, constituting an undersized prominentia cerebellaris. On the left, a shallow sulcus vena occipitalis externa is observed (Fig. 3c).
The foramen rami occipitalis arteria ophtalmica is open at both sides of the prominentia cerebellaris. The condylus occipitalis is kidney shaped and sub-rounded, with a slightly marked incisura medialis condylaris. The processus paroccipitalis is caudally extended, condition that might be caused by dorsal compression (Fig. 3 and Fig. 7). The processus zygomaticus is robust and laterally expanded. The cavum typanicum area is badly preserved, including the cotylae quadratica, which are either not preserved or impossible to visualize (Fig. 3f).
Openings of the n. hypoglossus are difficult to determine because this area is badly preserved. However, five openings were attributed to this nerve, besides the larger foramen nervi vagi (Fig. 3c, d).
The lamina parasphenoidalis is rhomboidal and dorsally located with respect to the condylus occipitalis. The fossa subcondylaris is deep and divided in left and right sides by a small ridge perpendicular to the crista basilaris transversa (Fig. 3e, f). The lamina parasphenoidalis is largely expanded and carries strong structures (processus mediales parasphenoidales, p. laterales parasphenoidales, and tubercula basilaria).
The processus mediales parasphenoidales are bulky, and cover the lateral boundaries of the lamina parasphenoidalis. Each process forms a thick triangle, ventrally rounded and pointing laterally (Fig. 3e, f). These structures are ventrally more prominent than the condylus occipitalis, and become slender toward their cranial and caudal ends. Not completely separated from the latter, the processus laterales parasphenoidales are more cranio-medially located and remain in contact with the crista basilaris transversa. A pair of well-defined tubercula basilaria are in the caudalmost end of the lamina parasphenoidalis, between the condylus occipitalis and the processus mediales parasphenoidales (Fig. 3e, f).
Unlike in Anthropornis grandis, the lamina parasphenoidalis is rectangular in MLP 12-I-20-1, and the whole lateral edges of this lamina are covered by the processus laterales parasphenoidales, which are half-moon shaped and reach the condylus occipitalis caudally. In Icadyptes, the laminaparasphenoidalis is sub-circular and the processus laterales parasphenoidales are well developed and cranio-caudally expanded, as in A. grandis. The processus mediales parasphenoidales are missing in Icadyptes and MLP 12-I-20-1.
In MLP 12-I-20-1, the crista basilaris transversa is more defined than in A. grandis and clearly bounders the cranial end of the lamina parasphenoidalis. The fossa subcondylaris is deeper in A. grandis than in MLP 12-I-20-1. The condylus occipitalis is blunter and larger in A. grandis than in MLP 12-I-20-1.
The rostrum parasphenoidalis is broken, preserving only the caudalmost portion. Parts of the ossa maxillare and premaxillare are preserved as isolated fragments.
Braincase pneumaticity. A complex series of pneumatic sinuses (e.g., Dufeau, 2011, Elzanowski and Galton, 1991 and Witmer, 1997, and references therein) permeate the avian braincase, and a reduction of these sinuses characterizes many diving birds (Ksepka et al., 2012b and Tambussi et al., 2015). Stem fossil penguins have a general pattern for tympanic recess similar to that of extant penguins, except for Paraptenodytes (Miocene of Argentina) and MLP 12-I-20-1 (Eocene of Antarctica), which retain the contralateral connection between the paired rostral tympanic recesses called the interaural pathway (Proffitt et al., 2016 and Tambussi et al., 2015). This “extra” pneumaticity with respect to extant penguins — the ancestral state in birds except for penguins —, may have had resulted in a potential difference in sound localization in stem taxa (Calford and Piddington, 1988 and Proffitt et al., 2016). On the contrary, Anthropornis grandis has no relict of the interaural pathway, as shown in Fig. 8a. Also, the conspicuous recess described for MLP 12-I-20-1 as caudal tympanic recess by Tambussi et al. (2015, fig. 7.2) is also present in A. grandis (Fig. 8b). The absence of the interaural pathway and the reduction of intracranial pneumaticity in general — regarding other Antarctic fossil penguins — seem to be a derived condition, and A. grandis exhibits an intermediate situation.
Quadratum. Left and right quadrates are preserved. The left one preserves the condylus of articulation with the mandible and part of the corpus, whereas the processus oticum and orbitalis are missing. The right one is more complete, missing only the processus orbitalis (Fig. 3b). Therefore, the tuberculum subcapitulare, place of origin of m. adductor mandibulae externus pars profunda, is not observed. The quadrates are robust, apneumatic, and massive. They present a sturdy corpus ossi quadrati and bulky condylus caudalis, lateralis, and medialis. The processus oticum is wide. The condylus medialis is ellipsoidal and proportionally larger than in extant forms, and the processus mandibularis diverges far from the corpus ossi quadrati. The condylus caudalis is less extended and closer to the corpus than in extant penguins (e.g., Aptenodytes forsteri). The condylus pterygoideus is well developed and rounded. The right quadratum preserves the tip of the processus oticum, where a sub-rounded capitulum oticum and the capitulum squamosum are separated from each other by a shallow and wide incisura intercapitularis. Both capitula have a tiny and sharp projection pointing ventrally toward the middle part of the corpus ossi quadrati. Two foramina pneumatica open at the base of the processus orbitalis. The proportional length between the corpus ossi quadrati and the processus orbitalis cannot be calculated in A. grandis because the latter is broken in both sides.
The quadratum of A. grandis is dorso-ventrally more extended than that of Icadyptes and Kairuku. In Icadyptes, condyla are bulkier and more rounded, and sturdier than in extant penguins, Kairuku and A. grandis. The incisura intercapitularis is similarly developed in A. grandis and Icadyptes. Although most of the processus orbitalis is broken, its base is widened, as in Kairuku and differs from the slender processus in living forms.
Mandible.Rami mandibulae are slender and extremely elongated (Fig. 4a). The ramus mandibulae are incomplete, missing the processus coronoideus and the angulus mandibulae.
The pars symphysialis is similar in extension to that one of Icadyptes, Inkayacu,Muriwaimanu, and Sequiwaimanu; it is convex ventrally and barely concave dorsally. In Kairuku, the tip of the mandible is completely different, for it has a moderate ventral deflection, whereas it is straight in all the other compared Paleogene mandibles.
The cristae tomiale are sharper at the cranialmost part. In the medial aspect, a deep sulcus runs along the ramus to its half, where it becomes shallower and disappears leaving a flat surface. The ramus mandibulae presents a lingual depression in Anthropornis grandis, feature previously described in Perudyptes (Ksepka and Clarke, 2010). However, in the latter, the ramus is caudally higher than in A. grandis.
Although the articular area is partially covered by sediment, each cotyla can be clearly distinguished. The cotyla medialis is more ventrally located than the others and it bends medially. The processus medialis mandibulae extends medio-caudally as a triangular projection. The facies articularis parasphenoidalis is large and similarly expanded to that one in Inkayacu. The cotyla caudalis is separated from the cotyla lateralis by a wide fossa. Cotylae lateralis and caudalis of the mandible are larger and stronger in A. grandis than in Inkayacu. Only the cotyla medialis presents a similar development in both species.
The crista transversa fossae is straight and it limits a wide sulcus intercotylaris. The crista intercotylaris is rounded and positioned near to the cotyla medialis. In A. grandis, a strong and rounded process immediately anterior to the cotyla medialis is present (Fig. 3g). Due to its position, it might be an exacerbated tuberculum pseudotemporale (this tubercle was mistaken in previous contributions for one that lies closer to the ventral margin of the mandible, which might be a tuberculum for insertion of the m. pterygoideus, see, e.g., Haidr and Acosta Hospitaleche, 2017a), with a degree of robustness not seen in any of the Eocene Antarctic ramus mandibulae. In caudal view, the posterior end of the mandible is heart-shaped, with a perfectly rounded foramen pneumaticum located in the dorso-medial corner of the fossa caudalis. The foramen pneumaticum is rounded and larger in Anthropornis grandis than in MLP 13-XI-28-199 and in MLP 96-I-6-48.
Costae. Several fragments of ribs are preserved, two of them with a quadrangular processus uncinatum fused. Cross-section of the ribs are sub-triangular. The sutura costouncinatum appears only on the facies lateralis (Fig. 4b, c).
Vertebrae. Several fragments including cervical and thoracic zygaphophysis caudalis are present. Unfortunately, the fragmentary state does not allow the comparison of any relevant feature.
Pygostyle. The basis pygostyli corresponding to the first fused caudal vertebra is preserved together with the basal part of the lamina pygostyli (Fig. 4d, e). The incompleteness of this element does not allow reliable comparisons.
Tarsometatarsus. It is robust and large, with a conspicuous eminentia intercotylaris aligned with the third metatarsal, but slightly medially inclined. Accompanying this asymmetry, the cotyla medialis is more dorsally placed than the cotyla lateralis, which is also smaller (Fig. 5c).
Metatarsals III and IV are parallel along all their extension (Fig. 5a, c), whereas metatarsal II diverges medially (Fig. 5a, c) and trochlea metatarsi II is more plantarly located (Fig. 5d). Trochlea metatarsi II is the smallest and its trochlear edges are weak. The central trochlea metatarsi III is laterally and medially rounded, the most robust, and the most distally projected. Its trochlear edges are the strongest among all the trochleae, particularly compared to the medial one. These edges slightly diverge distally and plantarly. Trochlea metatarsi IV is the less caudally projected and its trochlear edges are less pronounced than those of trochlea metatarsi III (Fig. 5a, c). Due to the lateral leaning of the edges, the medial edge is located at the same level than that of trochlea metatarsi III, whereas the lateral edge is lower and plantarly displaced (Fig. 5d).
As typical of the genus, the foramen vasculare proximale laterale is considerably smaller and more proximal than its medial counterpart (Fig. 5c). In Palaeeudyptes, the other genus of giant penguins from Antarctica, the lateral foramen is large and the medial foramen reduced or absent. Both foramina are separated by a wide third metatarsal III. As described in the revised diagnosis of Myrcha et al. (2002), the proximal end of metatarsal III is clearly lowered in relation to metatarsalia II and IV. Tuberositas musculi tibialis cranialis is wide, although not well marked.
The crista medialis hypotarsi is divided into two short crests, which are more separated from the stronger and wider crista lateralis hypotarsi. The last one is aligned with the second metatarsal and extends further distally, broadening at its distal end (Fig. 5a, b).
Phalanges. The phalanges I, II and the ungual of digiti II, phalanges I and II of digiti III, and phalanges I and II of digiti IV are preserved (Fig. 5c). The first phalanx of digit II is preserved in two pieces. The facies dorsalis is barely convex, whereas the facies plantaris is concave. The tuberculum extensorius is high and leans toward the metatarsal III. The tuberculum flexorium extends proximally just a little less than the t. extensorius. The cotyla articularis is represented by two symmetrical facets separated by a weak vertical rib extended between both tubercula. The dorsal extension of the fovea subtrochlearis is scarce and continues distally with the trochlea articularis. The caput phalangis is sub-rounded and the fovea ligamentariscollateralis is just marked. The entire corpus of the second phalanx of digit II is dorso-plantarly more depressed than the first phalanx. The phalangis ungualis surface is completely weathered, and most of its features are erased. Its general morphology represents a cone with an almost flat plantar surface (the tuberculum flexorium forms a blunt protuberance that covers the half of this face). The caput phalangis has a slight inclination toward the digit III.
The digit III is sturdier than the others, and the first phalanx is particularly massive. The hourglass shape is a little more accentuated, although the proximal end is wider and higher than the distal tip. The tuberculum flexorium is very conspicuous, and it indicates the limit between both cotylae in the proximal face. The medial cotyla articularis is wider than the lateral one. All the other features described before are very similar to this phalanx.
Finally, the first phalanx of digit IV is intermediate in size between the proximal phalanges of digits II and III. The tuberculum extensorius leans medially, and the lateral side of the corpus is straighter than the medial side. The cotylae articularis are deep and symmetrical. The second phalanx follows a similar morphology, although it is dorso-plantarly more depressed.
The cranial endocast of Anthropornis grandis is complete, but deformed. The main deformation forces were dorso-latero-ventral, resulting in a somehow more “flatten” cranial endocast. However, some structures, such as the cisura interhemispherica (deep straight longitudinal groove separating the cerebral hemispheres medially), both wulsts or sagittal eminences, the cerebellum, partial cerebral hemispheres, the optic lobe, partial pituitary, and right inner ear, can be observed (particularly in dorsal and right lateral views). The volume of the endocranial cast is difficult to determine due to the high degree of deformation. The size of the cerebral hemispheres (48.4 mm in length and 45.5–51.0 mm in width) and wulst (45.5 mm in length and 17.5 mm maximum in width) suggests a volume similar to that of other Antarctic stem penguins, calculated in a range of 27.3–28.8 cm3 (see Tambussi et al., 2015, tables 1 and 3).
The main axis of the cranial endocast is parallel to the main axis of the skull, as in all crown penguins, fossil stem penguins such as Paraptenodytes antarcticus, and Gavia (Ksepka et al., 2012b, Paulina-Carabajal et al., 2015, Proffitt et al., 2016 and Tambussi et al., 2015). There is a weak lateral expansion of the cerebral hemispheres, a noticeable feature described also for MLP 12-I-20-1 and MLP 12-I-20-2 by Tambussi et al. (2015), which results in a less “heart-shaped” outline in dorsal view (Fig. 9a, b). The length of the cerebral hemispheres is 48.4 mm, whereas the preserved width is 45.5 mm. These measurements are similar to those of MLP 84-II-1-10 (51.5 mm long, 47.5 mm width; Tambussi et al., 2015), being, in turn, larger than the other Antarctic fossil and extant penguins. In Anthropornis grandis, the fissura interhemispherica is wide and deep, particularly anteriorly, as in other Antarctic penguins (Tambussi et al., 2015), the Miocene Paraptenodytes and the Paleocene specimen CM 2013.27.1 described by Proffitt et al. (2016). This trait is not shared with crown penguins.
In Anthropornisgrandis, there is a well-defined wulst, as in living forms (e.g., Aptenodytes, Spheniscus), and Paraptenodytes and other Eocene Antarctic forms among fossil taxa (Bee de Speroni and Pirlot, 1987, Ksepka et al., 2012b and Tambussi et al., 2015), and opposed to the Paleocene penguin from New Zealand (specimen CM 2013.27.1 described by Proffitt et al., 2016), whose wulsts are markedly smaller. However, as stated for other Eocene Antarctic penguins, this structure is less dorsally expanded than in extant forms. In A. grandis, the wulst is elongated, occupying approximately the complete length of the cerebral hemisphere, extending almost to the level of the cerebellum. In extant penguins and Paraptenodytes, the wulst is more projected and extends all the way to the level of the cerebellum in dorsal view (Tambussi et al., 2015). In A. grandis, both wulsts are sub-elliptically shaped in dorsal view, with lateral margins almost parallel (not curved). The clear separation between the wulst and the rest of the cerebral hemisphere indicates that there was a conspicuous vallecula. Each wulst is not expanded posteriorly (as is in Pygosceliscalderensis, Paulina-Carabajal et al., 2015).
The cerebellum is globose, largely exposed in dorsal view and clearly differentiated from the wulsts. The cerebellum is similar to that described for MLP 12-I-20-1, having a relatively blunt caudal margin and a width greater than that of the medulla oblongata. In extant penguins and in Paraptenodytes, the cerebellum sharply tapers its caudal margin and is narrower than the medulla oblongata (Tambussi et al., 2015). The floccular process of the cerebellum (auricula cerebelli) is a prominent tongue-shaped structure (at least 8.9 mm in length), posterolaterally projected and closely related to the anterior semicircular canal of the inner ear (Fig. 9d). Endocranially, the auricula cerebelli is hosted within a large recess, the fossae auriculae cerebelli. The size and orientation of the flocculus are similar to what has been described for other Eocene Antarctic penguins by Tambussi et al. (2015), particularly MLP 12-I-20-1 and MLP 12-I-20-2, whereas MLP 84-II-1-10 has the longest flocculus (12.3 mm long).
The olfactory tracts are markedly short and the olfactory bulbs are relatively small (approximately 7.5 mm in length) and poorly differentiated from the forebrain, as in other penguins (Paulina-Carabajal et al., 2015 and Tambussi et al., 2015) (Fig. 9b). The olfactory ratio (OR) calculated for Anthropornisgrandis is about 15.5%, similar to that calculated for Spheniscusmagellanicus (16.4%) by Tambussi et al. (2015), and smaller than that calculated for the extinct MLP 12-I-20-1 (35%) and Paraptenodytes (24.5%).
The optic lobe is clearly visible in the right side of the cranial endocast. They are relatively well developed as in other living and extinct birds (Paulina-Carabajal et al., 2015, Walsh and Milner, 2011a and Walsh and Milner, 2011b) and postero-ventrally located regarding the cerebral hemispheres. Ksepka et al. (2012b) stated that the amount of variation in the form of the optic lobe within Spheniscidae is small compared with that between penguins and other groups.
The posterior region of the pituitary is preserved as a rounded structure at the base of the endocasts. As in the other Eocene (Tambussi et al., 2015) and Paleocene (Proffitt et al., 2016) penguins, the pituitary seems larger than in extant penguins. There are no observed neurovascular canals in the endocasts.
The inner ear is dorso-ventrally deformed, but several segments of the three semicircular canals, and the lagena, are preserved (Fig. 9a–d). It seems to exhibit the general avian morphology with large and slender semicircular canals and a long and slender lagena. It is likely that the orientation of the semicircular canals, particularly the anterior semicircular canal, is oblique relative to the anteroposterior axis of the skull, as in Pygoscelis. The semicircular canals are robust and the diameter of the tubes measures approximately 2.2 mm, similar to the robust semicircular canals of the Eocene Antarctic stem penguin MLP 12-I-20-1 (Tambussi et al., 2015), but larger than that in MLP 84-II-1-10. The anterior semicircular canal is the largest (Fig. 9d); it is oval and posteriorly inclined. It is not taller than the posterior semicircular canal. The angle between the asc and psc (in dorsal view) is slightly larger than 90 degrees, although, because of the deformation, this measurement cannot be confirmed. The lateral semicircular canal is complete and corresponds to an almost circular shape. As stated by Tambussi et al. (2015), the Antarctic fossil penguins have stouter (wider in diameter relative to overall labyrinth size) semicircular canals than in Paraptenodytes and the sampled extant penguins so far.
Muscular reconstruction
The temporal region, the area aspera on the orbital region, as well as the processus coronoideus, and a processus anterior to the articulation in the ramus mandibulae are conserved. They are all indicative of the development of some of the cranial muscles.
The m. adductor mandibulae pars rostralis would have had its surface of origin in a very robust and laterally extended proc. postorbitalis. Caudally, the stout proc. zygomaticus, which has a similar shape in extant forms, would have been its other point of origin. The plausible insertion of m. adductor mandibulae pars rostralis on the proc. coronoideus is not preserved.
The m. adductor mandibulae pars ventralis (Fig. 10a, b, c) origins on the fossa temporalis in extant forms. In A. grandis, this fossa is wide, deep, and reaches the sagittal plane forming a crista sagittalis (Fig. 3a). Spheniscus is the only extant penguin that develops a sagittal crest that is, however, significantly shorter than that of the Eocene forms. The fossa is larger and deeper than in living penguins, except again maybe in Spheniscus. The proc. orbitalis, which is the lateral limit of its development in extant forms is, as described above, more laterally extended than in any extant species. All of these features would be indicative of a robust and well-developed muscle with a much larger cross-sectional area than in any living species.
The m. depressor mandibulae attaches on the crista nuchalis transversa (Fig. 10a, b, c) in extant forms. This crista is thicker in Anthropornis than in modern species, and would have constituted a strong line of attachment for this muscle. The m. depressor mandibulae inserts on the processus retroarticularis (Fig. 10c), which in Anthropornis grandis is triangular, short, and wide, constituting a large area of origin. However, in comparison with extant forms, this area of insertion is smaller in A. grandis. Its lever arm would have been also shorter compared with extant forms such as Spheniscus, whose processus retroarticularis extends farther posteriorly. The m. pseudotemporalis superficialis, whose place of origin is on the area muscularis aspera on the os laterosphenoidale in living penguins, is preserved on the specimen (Fig. 10d). Its origin would have extended onto a fossa that lies underneath the proc. postorbitalis, condition not present in extant forms. The putative tuberculum pseudotemporalis, insertion point of m. pseudotemporalis superficialis (Fig. 10f) might indicate the presence of a strong tendon making contact, strengthening the fleshy insertion of the muscle in the ramus mandibulae. In extant forms, this tuberculum is more delicate and sometimes hard to identify. Regarding the m. protractor quadrati, its origin in the area muscularis aspera of the os laterosphenoidale, immediately ventral to that of the m. pseudotemporalis superficialis (Fig. 10b, d), presents a deep fossa that is clearly visible in the right side. This area of origin suggests that this muscle would have also been well developed.
Both palatines and pterygoids are missing, which is in detriment of a consistent reconstruction of the m. pterygoideus. The tuberculum in which the tendon of the m. pterygoideus Part M (sensuZusi, 1975) inserts (Fig. 10f, g) is much more robust than in extant forms, but less extended than in other Eocene mandibles such as MLP 92-II-2-115a, MLP 96-I-6-48, 13-XI-28-199, and 94-III-15-16a; these observations might be biased by the preservation state of the artefacts. The processus medialis of the mandible is well developed (area of attachment of the other parts of this muscle). It is very different to that of living forms, lacking the characteristic hook and extending caudally as opposed to the lateral development in extant forms. Because of these differences, the degree of development of this muscle cannot be certainly affirmed. The m. adductor mandibulae externus pars profunda, m. pseudotemporalis profundus, and the m. protractor pterygoidei could not be reconstructed, because both origin and insertion area lacking, or were unidentified.
Discussion and conclusions
MLP 14-XI-27-84 is the first Antarctic skull associated with postcranial elements and, therefore, the first opportunity to make an accurate systematic assignment based on the tarsometatarsus. This specimen, the largest penguin cranium discovered so far, is classified as Anthropornisgrandis, a giant species that lived in Antarctica during the Eocene. However, according to our knowledge of the giant forms, larger skulls are yet to be discovered.
The beak of this specimen (according to the mandible length) is elongated and slender as in other stem penguins, and the neurocranium is proportionally smaller than that of the extant species. Dimensions of the condylus occipitalis and the foramen magnum, however, are congruent with a large cranium and neck. Indeed, these proportions are widely extended among large and giant penguins during the Paleogene.
Secondary support of the mandible
The contact between the processus medialis parasphenoidalis (in the lamina parasphenoidalis) and the facies articularis parasphenoidalis (in the processus medialis mandibulae) works as a supplementary bony support of the mandible (Bock, 1960). This articulation is medial to the quadrate-articular joint and exists in many diverse groups of birds (e.g., Macronectes), according to our own observations. On the contrary, this abutment is absent in many other birds (e.g., Oceanites). In Anthropornisgrandis, this structure is unusually robust. It was assumed that this medial brace of the mandible serves to prevent the disarticulation of the mandible from the cranium (Bock, 1960). For that reason, it would be only necessary in birds that catch their prey or break the food by rapid movements of the head. If disrupting forces are strong enough, the medial brace of the mandible could compensate a deficient primary articulation with the condyla of the quadrate. It could be summarized that the functionality of this medial brace is determined by the kind of movements to which the joint is subject, the magnitude of the disruptive forces (during feeding), and the efficiency of the quadrate-articular joint (given by the morphology of the quadrate condyla and the mandibular cotylae).
MLP 14-XI-27-84 is similar to morpho-type 2 of Haidr and Acosta Hospitaleche (2012) because the cotyla lateralis is clearly separated from the cotyla caudalis. This morphology gives a greater stability and firmness to the mandibular articulation during sudden movements. Contrarily, in the morpho-type 1, the cotyla lateralis is merged with the cotyla caudalis. This condition was described for crustacivores (Haidr and Acosta Hospitaleche, 2012).
Although the quadrate in A. grandis is robust and each condylus is perfectly defined and separated from each other, they might be not rounded and bulky enough to resist the destabilizing forces. The great development of the processus medialis parasphenoidalis and of the facies articularis mandibulae could properly function as a reinforcing to prevent disarticulation during the capture and handling of prey.
In all birds, two muscles attach on the processus medialismandibulae: the m. depressor mandibulae inserts along the posterior edge, and the m. pterygoideus attaches on its anterior surface. Consequently, the enlargement of this surface would benefit the strength of any of these muscles. As it was pointed before, the additional bony support given by this process would be a preadaptation of the mentioned process (Bock, 1959 and Bock, 1960).
The significance of the fossa glandulae nasalis morphology
All the known fossil and living penguins develop a fossa glandulae nasale for the location of the salt gland. However, the shape and the development degree of the fossa glandula nasale are not clear indicators of the salt gland size (Ibañez, 2009). The supraorbital margin boundering laterally the fossa glandulae nasale is present in Pygoscelis, Eudyptes, and Megadyptes, penguins with a broad diet range. On the contrary, the piscivorous Spheniscus, Eudyptula, and Aptenodytes have the sulcus laterally opened, and the salt gland partially resting on the orbit (Ibañez, 2009). A similar condition appears in all the Paleogene taxa (including Anthropornis MLP 14-XI-27-84).
Muscular reconstruction: the cranio-cervical muscles
All the Eocene Antarctic crania (MLP 12-XI-1-1, MLP 12-XI-20-1, MLP 84-II-1-10, and UCMP 321265) present a wide fossa temporalis with a deep posterior portion, indicative of a large m. adductor mandibulae externus; the situation of Anthropornis grandis is not different. This large and deep area of insertion is comparable with that of other Eocene forms such as Perudyptes devriesi, Icadyptes salasi (Clarke et al., 2007), and also to Waimanu tuatahi (Proffit et al., 2016), as well as of geologically younger penguins such as Paraptenodytes antarcticus (Ksepka et al., 2012b), and Spheniscus (Stucchi et al., 2003).
Another muscle involved in the closing of the jaws is the m. pseudotemporalis superficialis (Fig. 10b), which place of both origin and insertion are more robust and well developed than in any of the other Antarctic Eocene forms (of the preserved material) or any extant forms. The biting force would have been powerful, with strong muscles that had large cross-sectional muscle areas. The situation of the m. depressor mandibulae (Fig. 10a) would have been different, having similar size with extant forms and a lever arm that would have been relatively short, a condition similar to Pygoscelis.
The origin point of the m. protractor quadrati indicates a conspicuous muscle that was contained in a much deeper fossa than that found in MLP 12-I-20-1 or extant species. This muscle moves the quadrate, lifting the upper beak (Zusi, 1975).
The presence and the development degree of the tubercula basilaria in Anthropornis grandis draw attention, especially because they are absent or inconspicuous in extant penguins (they may coincide with the processus mediales parasphenoidales according to Baumel et al., 1993). After examination of several specimens of modern penguins, we can establish that these structures are slightly developed (e.g., Aptenodytes forsteri) or simply lacking (e.g., Spheniscus magellanicus). The presence and development of these tubercles might be correlated with a functional demand of the neck muscles to efficiently support the skull. The tubercula basilaria appear in taxa not phylogenetically related, such as Stercorarius maccormicki (Charadriiformes), Pygoscelis adeliae (Sphenisciformes), and Cathartes aura (Cathartiformes), among others. These structures are the attachment area of cranio-cervical m. rectus capitis dorsalis and are typically more developed in long-skulled birds (Baumel et al., 1993), like the noticeably elongated skull of the specimen Anthropornis grandis here described. But, what does it means in functional terms? The M. rectus capitis dorsalis is, in fact, a series of discrete slips (converging towards the insertion) that originate from the anterolateral surface of the cervical vertebra C1, and transverse processes of cervical vertebrae C1–C6. This muscle inserts on the tubercula basilaria, if developed, or just in a rugose scar of the lamina parasphenoidalis. The m. rectus capitis dorsalis is responsible for head ventro-flexion relative to the neck (Snively and Russell, 2007a).
As it was suggested for dinosaurs (Snively and Russell, 2007b), a large insertion area (represented here by tubercula basilaria) may indicate a large cross-sectional area for m. rectus capitis dorsalis and a relatively more forceful contraction for the ventral and ventrolateral flexion of the head. This is consistent with the morpho-functional requirements that a predator like Anthropornis grandis would have had during capturing and handling its prey.
The interdependence of numerous adaptations of the head and neck was already noticed by Zusi (1962), who proposed the treatment of their totality as an adaptive complex for feeding. A set of characters observed in many penguins can be understood as adaptations for feeding specializations, e.g., innervation of mandible (for detecting prey), hypertrophy of jaw, curvature of the rostrum, and increment of kineticism (for striking prey), and a well-developed m. adductor mandibulae (for holding prey). However, some other features observed in Anthropornis grandis, such as the broad shape of the quadrate condyles and the increased contact of the mandibular processes and the lamina parasphenoidalis, are clearly developed for stabilization of jaws and head.
The development of a wide and deep fossa temporalis, a small neurocranium, a robust and large condylus occipitalis, and a long and spear-like beak, are among the most conspicuous features. It has been proposed that the Paleogene penguins would have caught large prey by spearing (Ksepka and Bertelli, 2006 and Olson, 1985), similar to the strategy of Podiceps (Chávez Hoffmeister, 2018). In this regard, whilst extant penguins depend on a powerful biting force, it was suggested that it would not be as important in Paleocene species (Chávez Hoffmeister, 2018). However, Anthropornis possess a wide and deep area of origin for the m. adductor mandibulae externus, which is the main muscle involved in the biting force. A similar condition appears in Podiceps major, Aechmophorus, and other podicipediforms. This is consistent with the development of voluminous and strong muscles. Whether by harpooning or biting, Paleocene penguins would still be in the need of strong forces for catching their prey.
Paleoneurology
The novel information on the cranial endocast of Anthropornis grandis reveals the relative proportions and orientation of the telencephalon, which resemble those of other Miocene Antarctic stem penguins (Tambussi et al., 2015) and the Paleocene penguin from New Zealand (Proffitt et al., 2016). This includes the poor lateral expansion of the cerebral hemisphere, the relative large development of the wulst, and the overlap between the cerebral hemispheres and the cerebellum in lateral view (Proffitt et al., 2016 and references therein).
As in the other studied Antarctic fossil penguins and the Paleocene penguin from New Zealand (Proffitt et al., 2016), the development of many structures of the cranial endocasts, such as the wulst and the flocculus of the cerebellum, is similar in extant penguins and in volant birds (Ksepka et al., 2012b); this is, in turn, consistent with the need for penguins to maneuver in complex three-dimensional space (Tambussi et al., 2015). Prominent flocculi seem to be a general feature for diving taxa. Given that most Eocene penguins were piscivorous (Acosta Hospitaleche, 2013), differences are not evidenced in the size of the flocculus of the cerebellum, nor in the inner ear, when compared to Pygoscelis (crustacivores forms) Spheniscus or Aptenodytes (piscivorous forms) — and therefore this structure is not helpful to better understand the paleobiology of Anthropornis. Also, the lack of cerebellar fold impressions on the endocasts is a trait shared by all wing-propelled diving birds (see Tambussi et al., 2015, and discussion therein).
In the case of Anthorpornis, the relative large size of the semicircular canals could be associated with diving skills, whereas the relatively more robust canals may indicate a slight difference in maneuverability when compared to extant penguins. The lagena is elongate and conical, not distally expanded, as described for other extinct and extant penguins (e.g., Paulina-Carabajal et al., 2015 and Tambussi et al., 2015). Regarding the olfaction, Tambussi et al. (2015) calculated the olfactory ratios for MLP 12-I-20-1, which had the greatest olfactory ratio for any sampled penguin, suggesting that higher levels of olfactory sensitivity may have characterized basal penguins compared with extant taxa. Although the relative importance of olfaction has not been quantified in penguins yet, the higher olfactory ratio in MLP 12-I-20-1 has been interpreted by the mentioned authors as a primitive trait. The olfactory bulbs are not clearly visible in the CT scans of A. grandis; yet, the estimated size and shape (see Fig. 9) suggest an olfactory ratio markedly similar to that of MLP 12-I-20-1.
Although the transitions of the neuroanatomy associated with the evolution of flight have been subject to extensive research, the shift to a diving ecology is less understood, and the correlation between diving and brain morphology remains poorly understood (Tambussi et al., 2015 and references therein). While there are many similarities in braincase and endocranial anatomy between Anthropornisgrandis, MLP 12-I-20-1, and MLP 84-II-1-10, the former exhibits a lesser degree of cranial pneumaticity and lacks the interaural connection. It is unclear for us if this difference responds to inter- or intraspecific variability.
Acknowledgments
Antarctic field trip was completely funded by Dirección Nacional del Antártico and the Instituto Antártico Argentino, and logistic support in the field was provided by Fuerza Aérea Argentina. Partial help was provided by the Consejo Nacional de Investigaciones Científicas y Tecnológicas and Universidad Nacional de La Plata (N838). Leonel Acosta (Museo de La Plata) prepared the fossil, María Bernarda Epele and Mariano Cipollone (Y-TEC) made Micro-CT scans, and Agustín Ruella together with Nicolás Bauzá (Museo de La Plata) performed the retro-deformation analyses. CAH is particularly grateful to Oceanwide Expeditions, Vlissingen (NL) for financial support. This project was partially funded by PICT-E 0080 (to División Paleontología Vertebrados, MLP), UNLP N838 (to CAH), and PICT-2016-0481 (to APC). María Florencia Sosa improved the English grammar. Three anonymous reviewers made important comments that improved our manuscript.
Acosta HospitalecheC.New crania from Seymour Island (Antarctica) shed light on anatomy of Eocene penguinsPol. Polar Res.342013397–412Acosta HospitalecheC.New penguin giant bones from Antarctica: systematic and paleobiological significanceC. R. Palevol132014555–560Acosta HospitalecheC.Paleobiological remarks on a new partial skeleton of the Eocene Antarctic penguin Palaeeudyptes klekowskiiAmeghiniana532016269–281Acosta HospitalecheC.Di CarloU.The coracoids in functional and morphological studies of penguins (Aves, Spheniscidae) of the Eocene of AntarcticaRiv. Ital. Paleontol. S116201023–34Acosta HospitalecheC.HaidrN.Penguin cranial remains from the Eocene La Meseta Formation, Isla Marambio (Seymour Island), Antarctic PeninsulaAntarct Sci.232011369–378Acosta HospitalecheC.RegueroM.First articulated skeleton of Palaeeudyptes gunnari from the late Eocene of Isla Marambio (Seymour Island), AntarcticaAntarct. Sci.222010289–298Acosta HospitalecheC.RegueroM.Palaeeudyptes klekowskii, the best preserved penguin skeleton from the Eocene–Oligocene of Antarctica: Taxonomic and evolutionary remarksGeobios47201477–85Acosta HospitalecheC.RegueroM.ScaranoA.Main pathways in the evolution of the Paleogene Antarctic SphenisciformesJ. S. Am. Earth Sci.432013101–111Acosta HospitalecheC.RegueroM.SantillanaS.Aprosdokitos mikrotero gen. et sp. nov., the tiniest Sphenisciformes that lived in Antarctica during the PaleogeneNeues Jahrb. Geol. Palaontol. Abh.283201725–34BaumelJ.J.KingA.S.BreazileJ.E.EvansH.E.Vanden BergeJ.C.Handbook of avian anatomy: Nomina Anatomica AviumPublication No. 231993Publications of the Nuttal Ornithological ClubCambridge, MassachusettsBee de SperoniN.PirlotP.Relative size of avian brain components of the Magellanic Penguin, the Greater Rhea, and the Tataupa TinamouMar. Ornithol.1519877–22BockW.J.Preadaptation and multiple evolutionary pathwaysEvolution131959194–211BockW.J.Secondary articulation of the avian mandibleAuk77196019–55CalfordM.B.PiddingtonR.W.Avian interaural canal enhances interaural delayJ Comp. Physiol.1621988503–510Chávez HoffmeisterM.Ocupación del morfoespacio y estrategias alimenticias en pingüinos y otras aves acuáticasActas VI Congreso Latinoamericano de Paleontología de Vertebrados, Villa de Leyva2018p. 71ClarkeJ.A.KsepkaD.T.StucchiM.UrbinaM.GianniniN.BertelliS.BoydC.A.Paleogene equatorial penguins challenge the proposed relationship between biogeography, diversity, and Cenozoic climate changeProc. Natl. Acad. Sci. USA104200711545–11550ClarkeJ.A.KsepkaD.T.Salas-GismondiR.AltamiranoA.J.ShawkeyM.D.D’AlbaL.BabyP.Fossil evidence for evolution of the shape and color of penguin feathersScience3302010954–957CracraftJ.The lacrimal-ectethmoid bone complex in birds: a single character analysisAmer. Midl. Naturalist.8021968316–359DufeauD.L.The evolution of cranial pneumaticity in Archosauria: patterns of paratympanic sinus development (Ph.D. dissertation)2011Ohio UniversityElzanowskiA.GaltonP.M.Braincase of Enaliornis, an Early Cretaceous bird from EnglandJ. Vert. Paleontol.11199190–107HaidrN.Acosta HospitalecheC.Feeding habits of Antarctic Eocene penguins from a morphofunctional perspectiveNeues Jahrb. Geol. Palaeontol. Abh.2632012125–131HaidrN.Acosta HospitalecheC.Fossil penguins beaks from the Eocene of Antarctica: new materials from La Meseta FormationVol. Especial Rev. MACN7201757–68HaidrN.Acosta HospitalecheC.A new penguin cranium from Antarctica and its implications for body size diversity during the EoceneNeues Jahrb. Geol. Palaeontol. Abh.2862017229–233HoussayeA.Pachyostosis” in aquatic amniotes: a reviewIntegr. Zool.42009325–340IbañezL.Anatomía de los surcos nasales y significado funcional de la glándula de la sal en aves marinas2009Tesis de licenciatura, Univ. CAECEBuenos Aires, ArgentinaJadwiszczakP.Eocene penguins of Seymour Island, Antarctica: the earliest record, taxonomic problems and some evolutionary considerationsPol. Polar Res.272006287–302JadwiszczakP.Partial limb skeleton of a “giant penguin” Anthropornis from the Eocene of Antarctic PeninsulaPol. Polar Res.332012259–274KsepkaD.T.BertelliS.Fossil penguin (Aves: Sphenisciformes) cranial material from the Eocene of Seymour Island (Antarctica)Hist. Biol.182006389–395KsepkaD.T.ClarkeJ.A.The basal penguin (Aves: Sphenisciformes) Perudyptes devriesi and a phylogenetic evaluation of the penguin fossil recordBull. Am. Mus. Nat. Hist.33720101–77KsepkaD.T.ClarkeJ.A.DeVriesT.J.UrbinaM.Osteology of Icadyptes salasi, a giant penguin from the Eocene of PeruJ. Anat.2132008131–147KsepkaD.T.FordyceR.E.AndoT.JonesC.M.New fossil penguins (Aves, Sphenisciformes) from the Oligocene of New Zealand reveal the skeletal plan of stem penguinsJ Vert. Paleontol.322012235–254KsepkaD.T.BalanoffS.WalshA.RevanA.HoA.Evolution of the brain and sensory organs in Sphenisciformes: new data from the stem penguin Paraptenodytes antarcticusZool. J. Linn. Soc.1662012202–219LivezeyB.C.ZusiR.L.Higher-order phylogenetics of modern Aves based on comparative anatomy: I. Methods and charactersBull. Carn. Mus. Nat. Hist.3720061–544MarplesB.J.Fossil penguins from the mid-Tertiary of Seymour Island, Falklands Islands DependenciesSurv. Sci. Rep.519531–15MayrG.De PietriV.L.LoveL.ManneringA.A.ScofieldR.P.A well-preserved new mid-Paleocene penguin (Aves, Sphenisciformes) from the Waipara Greensand in New ZealandJ. Vert. Paleontol.372017e1398169MeisterW.Histological structure of the long bones of penguinsAnat. Rec.431962377–386MontesM.NozalF.SantillanaS.MarenssiS.OliveroE.Mapa Geológico de la isla Marambio (Seymour) Escala 1: 20.0002013Serie Cartográfica Geocientífica AntárcticaMyrchaA.TaturA.Del ValleR.A new species of fossil penguin from Seymour IslandWest Antarctica. Alcheringa141990195–205MyrchaA.JadwiszczakP.TambussiC.P.NoriegaJ.I.GazdzickiA.TaturA.Del ValleR.Taxonomic revision of Eocene Antarctic penguins based on tarsometatarsal morphologyPol. Polar Res.2320025–46OlsonS.L.The fossil record of birdsAvian Biol.8198579–238Paulina-CarabajalA.Acosta HospitalecheC.Yury-YáñezR.E.Endocranial morphology of Pygoscelis calderensis (Aves, Spheniscidae) from the Neogene of Chile and remarks on brain morphology in modern PygoscelisHist. Biol.272015571–582ProffittJ.V.ClarkeJ.A.ScofieldR.P.Novel insights into early neuroanatomical evolution in penguins from the oldest described penguin brain endocastJ. Anat.2292016228–238SadlerP.M.Geometry and stratification of Uppermost Cretaceous and Paleogene units on Seymour Island, northern Antarctic Peninsula. Geology and Paleontology of Seymour Island, Antarctic PeninsulaGeol. Soc. Am. Mem.1691988303–320SharpeR.B.A review of recent attempts to classify birds1891Proceedings of the Second International Ornithological Congress, 90. BudapestSlackK.E.JonesC.M.AndoT.HarrisonG.L.FordyceR.E.ArnasonU.PennyD.Early penguin fossils, plus mitochondrial genomes, calibrate avian evolutionMol. Biol. Evol.2320061144–1155SnivelyE.RussellA.P.Functional morphology of neck musculature in the Tyrannosauridae (Dinosauria, Theropoda) as determined via a hierarchical inferential approachZool. J. Linn. Soc. Lond.1512007759–808SnivelyE.RussellA.P.Functional variation of neck muscles and their relation to feeding style in Tyrannosauridae and other large theropod dinosaursAnat. Rec.2902007934–957StucchiM.UrbinaM.GiraldoA.Una nueva especie de Spheniscidae del Mioceno tardío de la Formación PiscoPeru. Bull. Inst. Fr. Etud. Andin.322003361–375TambussiC.P.DegrangeF.J.KsepkaD.T.Endocranial anatomy of Antarctic Eocene stem penguins: implications for sensory system evolution in Sphenisciformes (Aves)J. Vert. Paleontol.352015e981635WalshS.MilnerA.Evolution of the avian brain and sensesDykeG.KeiserG.Living dinosaurs: the evolutionary history of modern birds2011John Wiley and Sons, LtdWest Sussex282–305WalshS.MilnerA.Halcyornis toliapicus (Aves: Lower Eocene, England) indicates advanced neuromorphology in Mesozoic NeornithesJ. Syst. Palaeontol.92011173–181WileyD.F.AmentaN.AlcantaraD.A.GhoshD.KilY.J.DelsonE.Harcourt-SmithW.RohlfF.J.St. JohnK.HamannB.MotaniR.FrostS.RosenbergerA.L.TallmanL.DisotellT.O’NeillR.Landmark Editor ver. 3.62007Institute for Data Analysis and Visualization (IDAV) and the University of CaliforniaDavishttp://www.idav.ucdavis.edu/research/EvoMorphWimanC.Über die alttertiären Vertebraten der Seymourinsel31905Wiss Ergeb Schwedischen Südpolar−Expd 1901–19031–37WimanC.Vorläufige Mitteilung über die alttertiären Vertebraten der SeymourinselBull. Geol. Inst. Uppsala.61905247–253WitmerL.M.The evolution of the antorbital cavity of archosaurs: a study in soft-tissue reconstruction in the fossil record with an analysis of the function of pneumaticityJ. Vert. Paleontol.1719971–76ZusiR.L.Structural adaptations of the head and neck in the Black Skimmer, Rynchops nigra LinnaeusPubl. Nuttall Ornithol. Club.319621–101ZusiR.L.An interpretation of skull structure in the penguins. The biology of penguins197559–84
Location map and stratigraphic column. a: Seymour Island, the star points the fossil locality IAA 4/12, where Anthropornis grandis MLP 14-XI-27-84 was collected; b: lithostratigraphic units showed in the map; c: general location of Seymour Island and of the Antarctic Peninsula; d: stratigraphic column of the Submeseta Formation; the star indicates the fossil locality IAA 4/12.
Carte de localisation et colonne stratigraphique. a : Seymour Island, l’étoile indique la localité fossilifère IAA 4/12, où Anthropornis grandis MLP 14-XI-27-84 a été collecté ; b : unités lithostratigraphiques indiquées sur la carte ; c : emplacement général de l’île Seymour et de la péninsule Antarctique ; d : colonne stratigraphique de la formation Submeseta ; l’étoile indique la localité du fossile IAA 4/12.
Associated material of Anthropornis grandis MLP 14-XI-27-84 before its complete extraction from the matrix (scale bar = 10 mm).
Matériau associé à Anthropornis grandis MLP 14-XI-27-84 avant son extraction complète de la matrice (barre d’échelle = 10 mm).
Skull of Anthropornis grandis MLP 14-XI-27-84. a: cranium in dorsal view (scale bar = 20 mm); b: right quadratum (scale bar = 10 mm); c: cranium in occipital view (scale bar = 10 mm), the circle indicates the magnified area in d; d: close-up of the area indicated in c (scale bar = 5 mm); e: schematic drawing of the lamina parashenoidalis and its main processes (same scale as in f); f: cranium in ventral view; g: cranium in lateral view (scale bar = 20 mm).
Crâne d’Anthropornis grandis MLP 14-XI-27-84. a : crâne en vue dorsale (barre d’échelle = 20 mm) ; b : quadratique droite (barre d’échelle = 10 mm) ; c : crâne en vue occipitale (barre d’échelle = 10 mm) ; le cercle indique la zone agrandie en d ; d : gros plan de la zone indiquée en c (barre d’échelle = 5 mm) ; e : dessin schématique de la lamina parashenoidalis et de ses processus principaux (même échelle qu’en f) ; f : crâne en vue ventrale ; g: crâne en vue latérale (barre d’échelle = 20 mm).
Mandible, rib and pygostyle of Anthropornis grandis MLP 14-XI-27-84. a: mandible in dorsal view, deformation of the mandible prevent a strict dorsal perspective; b: rib in lateral view; c: rib in medial view; d: pygostyle in cranial view; e: pygostyle in lateral view (scale bar = 20 mm).
Mandibule, côte et pygostyle d’Anthropornis grandis MLP 14-XI-27-84. a : mandibule en vue dorsale, la déformation de la mandibule empêche une perspective dorsale stricte ; b : côte en vue latérale ; c : côte en vue médiale ; d : pygostile en vue crânienne ; e : pygostile en vue latérale (barre d’échelle = 20 mm).
Left tarsometatarsus of Anthropornis grandis MLP 14-XI-27-84. a: caudal view; b: proximal view; c: cranial view; d: distal view (scale bar = 10 mm).
Tarsométatarse gauche d'Anthropornis grandis MLP 14-XI-27-84. a : vue caudale ; b : vue proximale ; c : vue crânienne ; d : vue distale (barre d’échelle = 10 mm).
Symmetrical points marked on the cranium of Anthropornis grandis MLP 14-XI-27-84 to perform the retro-deformation: a: caudo-ventral view; b: ventral view. The arrows indicate retro-deformation direction; the sagittal plane in black is established from the symmetrical points.
Points symétriques marqués sur le crâne d’Anthropornis grandis MLP 14-XI-27-84 pour effectuer la rétro-déformation. a : vue caudo-ventrale ; b : vue ventrale. Les flèches indiquent la direction de la rétro-déformation, le plan sagittal en noir s’établissant à partir des points symétriques.
Cranium of Anthropornis grandis MLP 14-XI-27-84 showing the original and the retro-deformed images. a: occipital view; b: ventral view; c: lateral view. The arrows indicate the change direction during the retro-deformation process.
Crâne d’Anthropornis grandis MLP 14-XI-27-84 montrant l’original écrasé et les images rétro-déformées. a : vue occipitale ; b : vue ventrale ; c : vue latérale. Les flèches indiquent le changement de direction du crâne déformé lors du processus de rétro-déformation.
CT scan slices of the skull of Anthropornis grandis MLP 14-XI-27-84 showing the lack of large pneumatic cavities. a: sagittal slice; b: coronal slice. Not to scale.
Scanner des coupes du crâne d’Anthropornis grandis MLP 14-XI-27-84 montrant l’absence de grandes cavités pneumatiques. a : tranche sagittale ; b : tranche coronale. Pas à l’échelle.
Digitally rendered skull and cranium endocast of Anthropornis grandis MLP 14-XI-27-84. The bone is semi-transparent to allow observation of internal structures. a,b: dorsal views; c,d: left lateral views (scale bar = 10 mm).
Endocast crânien et crâne numériquement rendus d’Anthropornis grandis MLP 14-XI-27-84. L’os est semi-transparent pour permettre l’observation des structures internes. a,b : vues dorsales ; c,d : vues latérales gauches (barre d’échelle = 10 mm).
Muscle attachment areas on the cranium of Anthropornis grandis MLP 14-XI-27-84 in a: left lateral view; b: right lateral view; c: dorsal view; d: right ventrolateral view; on the right caudal end of mandible in e: caudal view; f: medial view; g: ventral view.
Zones de fixation musculaire sur le crâne d’Anthropornis grandis MLP 14-XI-27-84 en a : vue latérale gauche ; b : vue latérale droite ; c : vue dorsale ; d : vue ventrolatérale droite ; à l’extrémité caudale droite de la mandibule: e : vue caudale ; f : vue médiale ; g : vue ventrale.
List of the cranial materials assigned to penguins from Seymour Island, Eocene of Antarctica.
Liste des matériaux crâniens attribués aux manchots de l’île Seymour, Éocène de l’Antarctique.
Table 1Repository numberMaterialReferenceIB/P IB-009912 cranial fragmentsMyrcha et al., 1990IB/P/B − 0167Partial craniumMyrcha et al., 1990IB/P/B − 0346Partial craniumJadwiszczak, 2006UCMP 321265Partial craniumKsepka and Bertelli, 2006UCMP 321208 UCMP 321223 UCMP 318657Cranial fragmentsKsepka and Bertelli, 2006MLP 84-II-1-10Partial craniumAcosta Hospitaleche and Haidr, 2011MLP 12-I-20-1Partial craniumAcosta Hospitaleche, 2013MLP 12-I-20-2Partial craniumAcosta Hospitaleche, 2013MLP 12-XII-28-8Partial craniumHaidr and Acosta Hospitaleche, 2017bMLP 14-XI-27-84Cranium/mandiblePresent contribution