A New Protocetid Whale (Cetacea: Archaeoceti) from the Late Middle Eocene of South Carolina

This paper describes the petrosal (periotic) and the inner ear of Herpetocetus sp., an archaic mysticete whale (Mysticeti, Mammalia) from the Yorktown Formation (Pliocene) of North Carolina, USA. Parsimony analysis of 28 petrosal characters of Herpetocetus sp. and 11 other cetacean taxa supports the monophyly of mysticetes and the division of odontocetes and mysticetes. The in-group taxa of this analysis are: Herpetocetus, Parietobalaena, Pelocetus, Balaenidae, Eschrichtius, and Balaenopteridae. Odontocetes and the archaeocete Zygorhiza were used as successive outgroups to root phylogenetic trees and to establish character polarities. Among the modern mysticetes, the Balaenopteridae (rorquals) and the Eschrichtiidae (gray whales) are more closely related to each other than either is to the Balaenidae (bowhead and right whales). Several Miocene “cetotheriid” mysticetes and balaenids share some resemblance in the petrosal, suggesting their affinities. Quantitative information of the inner ear of Herpetocetus sp. was obtained by serial sectioning and computer graphic reconstruction. Herpetocetus sp. is much less developed than odontocetes in the cochlear structures that are crucial for high frequency hearing. Some cochlear structures in this fossil mysticete resemble more closely the non-echolocating modern mysticetes than early fossil toothed whales, indicating a possible specialization in low frequency hearing. This suggests that the archaic mysticetes of the Miocene and Pliocene did not have high frequency hearing necessary for echolocation. Herpetocetus sp. is similar to modern mysticetes but different from odontocetes in the spherical shape of the vestibule.

The inferred transition from terrestrial hoofed mammal to fully aquatic cetacean has been intensively studied with fossil evidence. However, large sections of this remarkable evolutionary sequence are missing. Phylogenetic analysis of extant taxa may help to fill in some of these gaps. In this report, kappa-casein (exon 4) and beta-casein (exon 7) milk protein genes from cetaceans and other placental mammals were PCR-amplified, sequenced, and aligned to previously published sequences. Phylogenetic analyses of the casein data suggest that hippopotamid artiodactyls are more closely related to cetaceans than to other artiodactyls (even-toed hoofed mammals). An analysis of the nuclear casein sequences combined with published mitochondrial cytochrome b DNA sequences also supports the Cetacea/Hippopotamidae sister group. This affinity implies that some of the aquatic traits of cetaceans were derived in the common ancestor of Cetacea and Hippopotamidae. An extant "missing link" to Cetacea may have been overlooked by science since the description of the semiaquatic Hippopotamus in 1758. Paleontological information is grossly inconsistent with this hypothesis. If the casein phylogeny is accurate, large gaps in the fossil record as well as extensive morphological reversals and convergences must be acknowledged.

Recent phylogenetic analyses of DNA sequences suggest that cetaceans (whales) and hippopotamid artiodactyls (hippos) are extant sister taxa. Consequently, the shared aquatic specializations of these taxa may be synapomorphies. This molecular view is contradicted by paleontological data that overwhelmingly support a monophyletic Artiodactyla (even-toed ungulates) and a close relationship between Cetacea and extinct mesonychian ungulates. According to the fossil evidence, molecular, behavioral, and anatomical resemblances between hippos and whales are interpreted as convergences or primitive retentions. In this report, competing interpretations of whale origins are tested through phylogenetic analyses of the blood-clotting protein gene gamma-fibrinogen from cetaceans, artiodactyls, perissodactyls (odd-toed ungulates), and carnivores (cats, dogs, and kin). In combination with published DNA sequences, the gamma-fibrinogen data unambiguously support a hippo/whale clade and are inconsistent with the paleontological perspective. If the phylogeny favored by fossil evidence is accepted, the convergence at the DNA level between Cetacea and Hippopotamidae is remarkable in its distribution across three genetic loci: gamma-fibrinogen, the linked milk casein genes, and mitochondrial cytochrome b.