Lawrence G. Barnes

Outline of Eastern North Pacific Fossil Cetacean Assemblages

Barnes, Lawrence G. (Section of Vertebrate Paleontology, Natural History Museum of Los Angeles County, Los Angeles, California 90007) 1976. Outline of eastern North Pacific fossil cetacean assemblages. Syst. Zool. 25:321-343.-Geologic formations on the west coast of North America from Baja California, Mexico to central California, U.S.A., provide fossil evidence for a succession of Tertiary cetacean assemblages. Formations representing time periods of stage magnitude from early Miocene to Pleistocene age produce diverse aggregates of Cetacea containing fewer species than can now be found in latitudes corresponding to the Californian Province in the Pacific Ocean. Most fossil assemblages include a sperm whale, several dolphin-like taxa, and 3 or 4 baleen whales, although taxonomy is unstable, and many specimens cannot be assigned to named genera and species. New records of the earliest North Pacific occurrence of the families Balaenidae, Ziphiidae, Monodontidae, Phocoenidae, and Delphinidae (sensu stricto) date from the Miocene, and of the Eschrichtiidae from the Pleistocene. Squalodontidae are notable by their rarity. Early Miocene assemblages are dominated by Eurhinodelphidae, and late Tertiary ones by Stenodelphininae and primitive Balaenopteridae. On the generic level, there is a high degree of endemism among small Odontoceti, and a low degree among Physeteridae and Mysticeti. Collecting biases and deficiencies prevent recognition of antitropical distributions of fossil taxa, and paleoclimatology can not yet be inferred from the fossils. The status of the knowledge of fossil Cetacea in marine sedimentary deposits along the western border of North America is such that a summary and prospectus can now be formulated. A variety of Cenozoic fossil cetaceans are documented in the existing paleontologic literature, but these in no way reflect the real abundance of taxa or of specimens in the field or in museum collections. This article is in part, therefore, a summary of previously unrecorded specimens under study by the author (Barnes, 1972b). Mitchell (1966b) presented an earlier, more popularized overview of marine mammal evolution in the North Pacific Ocean. More fossil species are now represented by skulls and skeletons than previously, and some new interpretations of taxonomy and morphology are forthcoming in more detailed studies. These more complete specimens provide, in some cases, an opportunity for more precise determinations of phylogenetic and taxonomic relationships than was possible in the past for taxa founded upon fragmentary specimens. Kellogg (1923:38) has pointed out that comparisons among many fossil Cetacea have been hampered because of the lack of directly comparable parts of skeletons. Additionally, some taxa are described from skeletal parts having dubious taxonomic value. Cope (1868a, 1868b), Leidy (1851, 1869), and Deraniyagala (1969) have used isolated vertebrae and teeth, and Kellogg (1931) has used isolated periotics. In reality, both convergence and evolutionary conservativism may cause vertebrae and even teeth to be non-diagnostic in cetaceans. Even the isolated periotic has been proven, among Recent Cetacea, to sometimes be only generically and not specifically identifiable (Kasuya, 1973:72). This is particularly true for species of Kogia Gray, 1846 and for Delphinus bairdi Dall, 1873 and Stenella attenuata Gray, 1846 (see Kasuya, 1973:figs. 71-72). Recently, Ginsburg and Janvier (1971); and Hatai, Hayasaka, and Masuda (1963), have used isolated teeth and auditory bullae as holotypes. In light of such recent digressions in the quality of fossil cetacean systematics, I recommend that all subsequent cetacean holotypes include at least skull parts,

Paleoecology of whale-fall habitats from deep-water Oligocene rocks, Olympic Peninsula, Washington state

Fossil mollusks associated with eight Oligocene whales from the Makah and Pysht Formations on the northwestern part of the Olympic Peninsula, Washington, suggest that whale carcass sulfide production supported small numbers of some chemosymbiotic invertebrates as early as 30 million years ago. Thyasirid and modiolid bivalves usually dominate these very localized molluscan assemblages; lucinid and nuculanid bivalves, scaphandrid, naticid, and buccinid gastropods are rarely present; brachiopods were found once. These fossils include, tentatively, the first fossil record for the bivalve genus Idasola, and the first record of the bivalve Thyasira peruviana Olsson, outside of probable cold-seep deposits in South America. Strata surrounding the fossil whales contain low diversity megafaunas that include rare deep-water gastropods and bivalves, large isopods, and localized cold-seep communities. The whale-fall assemblages differ significantly because of the presence of Idasola? sp. and Thyasira peruviana? and the absence of vesicomyid bivalves. Vertebrate carcasses have probably not contributed significantly to the dispersal of cold-seep and hydrothermal vent invertebrates. Seep/vent communities were well established much earlier than the evolution of cetaceans, and seep/vent invertebrates have not been found with carcasses of other large vertebrates.

Origin of a widespread marine bonebed deposited during the middle Miocene Climatic Optimum

Bonebeds are vertebrate bioclast concentrations in beds that are local to basinal in extent. The middle Miocene Sharktooth Hill bonebed in the southeastern San Joaquin Basin of California is among the largest of such deposits, exposed over 15 km and containing a rich assemblage of marine vertebrates, with a mean density of ~200 specimens/m 2 . It ranks among the most widespread and richest bonebeds known, yet its genesis is poorly understood. Hypotheses for its origin and formation include mass death from shark predation, volcanic or red tide poisoning, accumulation from a calving ground for marine mammals, and condensed accumulation over a long period of time. Based on multiple kinds of evidence, we conclude that the bonebed formed over a protracted time interval of little to no net clastic sedimentation, coincident with a significant transgressive-regressive cycle between 16 and 15 Ma ago, during the middle Miocene Climatic Optimum (MMCO). Geochronological constraints bracket the duration of bonebed formation to no longer than 700 ka, indicating that time averaging is a critical consideration for paleoecological analyses of North Pacific Ocean biotic richness during the MMCO.

A NEW GENUS AND SPECIES OF LATE MIOCENE PONTOPORIID DOLPHIN (CETACEA: ODONTOCETI) FROM THE ST. MARYS FORMATION IN MARYLAND

Abstract Stenasodelphis russellae, a new genus and species of extinct dolphin in the odontocete family Pontoporiidae, is based on a partial cranium of Late Miocene age (Tortonian), circa 9 to 10 Ma, from the Little Cove Point Member of the St. Marys Formation, Maryland, USA. Pontoporiidae are amongst the smallest cetaceans, and this dolphin is one of the smallest reported pontoporiids, being probably less than 1.5 m long. It is also the second oldest named pontoporiid in the world. Stenasodelphis russellae shares with late Middle Miocene Brachydelphis mazeasi (eastern South Pacific) and the latest Miocene and Pliocene species of Parapontoporia (North Pacific) left-skew asymmetry of the cranial vertex. Thus, all of the earliest known Pontoporiidae have cranial asymmetry, in contrast to symmetrical crania in the Pliocene pontoporiids, Pontistes rectifrons and Pliopontos littoralis, and the Recent Franciscana, Pontoporia blainvillei. This suggests that cranial asymmetry may be the primitive character state among stem Pontoporiidae. Autapomorphies of Stenasodelphis russellae include small size, thick cranial bones, wide premaxillary sac fossae, a medial projection of each maxilla onto the lateral side of each nasal, and the highest part of the cranial vertex being formed by the nasals rather than the frontals.

A new name for the ‘Stanford skeleton’ of Paleoparadoxia (Mammalia, Desmostylia)

One of the most important and well-known fossil skeletons of the Miocene herbivorous, quadrupedal marine mammal Paleoparadoxia has been known informally for many years as the ‘Stanford skeleton,’ or the ‘Stanford Paleoparadoxia.’ Only recently has this specimen been formally referred to a named species of the genus. Unfortunately, that referral is tenuous, and it threatens nomenclatural stability and universality because it involves transferring a well-established name from one species to another. We propose to solve this problem by designating the ‘Stanford skeleton’ as the holotype of a new species bearing an unambiguous specific name. On 2 October 1964, during construction of the Stanford University Linear Accelerator Center near Menlo Park, California, a nearly complete, articulated skeleton of a large adult desmostylian was discovered. Its excavation, collection, and preparation were supervised by the late Charles A. Repenning, then of the United States Geological Survey office in Menlo Park (Fig. 1), who undertook its preliminary study, referred it to the genus Paleoparadoxia (Repenning, 1965; Repenning and Packard, 1990), and promoted the distribution and exhibition of replicas of it (see also Panofsky, 1998). The ‘Stanford skeleton’ (which lacks the cranium, the right dentary, and most of the dentition) has subsequently become widely known among marine mammal paleontologists, and it has been included in several comparative and functional studies. Replicas are exhibited in museums in several countries, and pictures of the skeleton in various poses have appeared in print (Repenning, 1965; Mitchell and Lipps, 1965; Romer, 1966:fig. 367 [after Repenning, 1965]; Inuzuka, 1982; Panofsky, 1998:1, fig. 56). Only recently, however, has a comprehensive, thoroughlyillustrated formal description of the skeleton been published (Inuzuka, 2005). The species nomenclature of the ‘Stanford skeleton’, nonetheless, remains in question. As was detailed by Inuzuka (2005), the type species of the genus Paleoparadoxia Reinhart, 1959, is P. tabatai (Tokunaga, 1939), the type material of which was a left m2 and a fragment of an upper tooth (possibly representing different individuals), found during construction of a tunnel between the towns of Sawane and Aikawa on Sado Island, Niigata Prefecture, Japan. Both of these teeth were destroyed during World War II. Subsequently, another Paleoparadoxia specimen from Honshu, Japan (known as the ‘Izumi skeleton,’ a subadult), was designated as the neotype of P. tabatai by Shikama (1957, 1966). This name, Paleoparadoxia tabatai, has for several decades been uniformly applied to the subadult ‘Izumi skeleton’ and to similar specimens until Inuzuka (2005) proposed the new species name P. media, having as its holotype the ‘Izumi skeleton.’ Inuzuka’s reason for this action was the recent rediscovery of another tooth (a left m3) which possibly represents the same individual as the lost left m2 of the original holotype of P. tabatai (the ‘Sawane specimen’). This rediscovery, as interpreted by Inuzuka (2005), automatically sets aside Shikama’s (1957, 1966) designation of the ‘Izumi skeleton’ as the neotype of Paleoparadoxia tabatai, and reinstates the ‘Sawane specimen’ as the holotype of P. tabatai (ICZN, 1999:Article 75.8). This rediscovered m3 that is possibly part of the type material of P. tabatai is considerably larger than the homologous tooth of the ‘Izumi skeleton,’ is geochronologically younger (16.5 Ma versus 18 Ma), and the two specimens most likely do actually represent different species. The root morphology of the rediscovered large left m3 of the supposed type material of P. tabatai (the ‘Sawane specimen’) is compatible with the vacant alveoli of the missing m3 in the left dentary of the ‘Stanford skeleton’ (which at approximately 14 Ma is younger geochronologically than the holotype of P. tabatai by about 2.5 Ma). Thus, it was solely on the basis of root morphology of the otherwise unknown m3 of the previously unnamed ‘Stanford skeleton’ that Inuzuka (2005) referred that specimen to the approximately 2.5 Ma older species, P. tabatai. A result of this decision by Inuzuka (2005) is that the species name P. tabatai, which for nearly half a century has been universally used for the ‘Izumi skeleton’ and others that have been deemed conspecific with it, has now been transferred to the much larger and geochronologically younger species that is represented by the ‘Stanford skeleton.’ This species transfer was done, moreover, on the tenuous basis of similarity in size and morphology between the roots of the m3 of the approximately 14 Ma ‘Stanford skeleton’ and of the m3 of the approximately 16.5 Ma ‘Sawane specimen,’ which is possibly the only surviving portion of the type material of P. tabatai. Our view is that even if the woefully incomplete ‘Sawane specimen’ (the holotype of P. tabatai) and the wonderfully complete ‘Stanford skeleton’ could ever be proven to be conspecific, the transfer by Inuzuka (2005) of the name P. tabatai from one species to another is seriously destabilizing to nomenclature, and will create a stumbling block for persons consulting desmostylian literature in the future. As the Code emphasizes (ICZN, 1999:General Recommendation 1), “it is of especial importance that a name should not be transferred to a taxon distinct from that to which it is generally applied.” Furthermore, it is now obvious that the family Paleoparadoxiidae was taxonomically diverse. The oldest and most primitive known paleoparadoxiids are species of the late Oligocene Behemotops from the Olympic Peninsula in Washington, U.S.A. (Domning et al., 1986; Ray et al., 1994; Barnes and Goedert, Corresponding author. Journal of Vertebrate Paleontology 27(3):748–751, September 2007

The Early Miocene Odontocete Araeodelphis natator Kellogg, 1957 (Cetacea; Platanistidae), from the Calvert Formation of Maryland, U.S.A.

ABSTRACT On the basis of an assigned specimen (USNM 526604, from the Plum Point Member of the Calvert Formation, Early Miocene, Maryland, U.S.A.), Araeodelphis natator Kellogg, 1957, is referred to the Platanistidae. A phylogenetic analysis identifies A. natator as the most stemward member of the family. By contrast, the extant river dolphin, Platanista gangetica (Platanistidae), is one of the most specialized odontocetes. Araeodelphis natator exhibits the following unique combination of characters: overall skull length (condylobasal length) estimated at about 50 cm; rostrum twice the length of the facial region; rostrum wider than deep throughout its entire length; approximately 50 teeth in each quadrant of rostrum; mesorostral canal closed dorsally through anterior half of rostrum by apposition of contralateral premaxillae; cranium with elevated orbits directed anterolaterally; maxillary crest (supraorbital process of frontal and overlapping maxilla) modestly thickened laterally and elevated above midline of skull; non-pneumatized supraorbital eminences; lobe of the pterygoid airsac sinus occupying orbital surface of frontal; zygomatic process compressed transversely; no postglenoid process; and glenoid facet faces medially. Araeodelphis provides new data about the definition and phylogenetic relationships of platanistids with other platanistoids, confirming the sister-group relationship with the extinct squalodelphinids and the ancestral platanistid skull morphology preceding the platanistine-pomatodelphinine split.