Sunday, January 24, 2016

2015 in review: Advances in marine mammal paleontology

Welcome to the fourth annual review of advances in marine mammal paleontology! Similar to 2014, there were just over 50 new papers. I've managed to read most of them in detail, and am sufficiently caught up in reading that I can take a break and read some of what's come out in 2016. I'll admit that I spent most of winter break relaxing, visiting museums, and doing fieldwork - and as a result at least 1/2 or 1/3 of this has been typed up since January 1st. My apologies for being about three weeks late. Also, you're welcome.

As usual, I'm sure I've missed something. If so, let me know, and if I'm not too sick and tired of working on this, I'll go ahead and add it in. 

In the 1980s Daryl Domning and Christian de Muizon reported a small sirenian rib from the Pisco Formation of Peru, one of the few records of dugongid sea cows from the Pacific coast of South America. Years later it was more precisely identified as a rib of the small sea cow Nanosiren, but considered to be strange in comparison to most other sea cows. This new study by Eli Amson and others took a histological slide from the rib and compared it with sirenians and the aquatic sloth Thalassocnus. Sea cows typically have an unremodeled layer of sheet-like cortical bone with an inner zone consisting of remodeled "secondary osteons"; Thalassocnus lacks this parallel-sheet like zone and instead the entire cross section consists of remodeled bone. The mystery rib matches the histological pattern of Thalassocnus, and is reidentified by the authors as an aquatic sloth rib. The authors point out that the most highly specialized aquatic sloths now do not overlap stratigraphically with sea cows in the Pisco Formation, suggesting that the sea cow niche was occupied by Thalassocnus after sirenians went extinct in the western south Atlantic.

This new study reports a fragmentary mysticete mandible from the Azores Islands off the coast of North Africa from Pleistocene strata. This mandible is not complete enough to identify to the family level, but is significant owing to its geographic location and strange bone modifications. It has a series of large pits on one surface, which appear to be anatomically genuine (e.g. ante-mortem rather than post-mortem) and thus not taphonomic in origin. The authors point out that relatively few fossil cetaceans have been reported from small oceanic islands, and hypothesize that such occurrences are likely underreported. However, it's worth noting that small islands are unlikely to have large sedimentary basins and that most small islands have Cenozoic strata formed as "bathtub ring" deposits where the abundance and ease of discovery for fossil cetaceans is likely going to be rather low in comparison to continental margin deposits. Then again, cetacean bones are not small (quite the opposite) so there's always a trade off.

Desmostylians are some of the most wonderfully bizarre of all marine mammals. Well known Miocene forms like Desmostylus, Paleoparadoxia, and the recently split-off Archaeoparadoxia and Neoparadoxia, are so derived that it's difficult to pick out exactly which group of plodding semiaquatic herbivores they belong to. Tradition dictates that they're tethytheres, most closely related to sirenians and proboscideans - but the possibility remains that they're perissodactyls. More fossils of early desmostylians are needed, and in 1986 Daryl Domning, Clayton Ray, and Malcolm McKenna published some late Oligocene specimens from Oregon that my hero Doug Emlong had collected - some mandibles and other elements, upon which they erected two species - Behemotops proteus and Behemotops emlongi - the "Behemoth face". None of the skull was known, but the more primitive teeth permitted the authors to link desmostylians with the anthracobunids, hippo-like relatives of early elephants. Later finds of Behemotops led the same authors to reevaluate the more fragmentary species B. emlongi and synonymize it with B. proteus. Still, no skull was known - until Brian Beatty and Thomas Cockburn reported this new specimen of Behemotops cf. proteus with most of a skull, some teeth, and quite a bit of the postcranial skeleton. The new specimen indicates that the rostrum was unusually elongate and narrow, similar to Cornwallius. The rather wide "shovel jaw" of B. emlongi is proportionally much wider than the rostrum of Behemotops cf. proteus, and so these authors erected the new genus Seuku to house the other species - Seuku emlongi. These Oligocene desmostylians indicate that several different desmostylians inhabited the same area at the same time - Cornwallius sookensis, Behemotops proteus, and Seuku emlongi - apparently coexisted in the Pacific Northwest, paralleling ecological diversity in sea cow assemblages.

This textbook is the best edition yet - the first edition came out in 1999, and the second edition came out in 2006. This text deals with most aspects of the biology and evolution of marine mammals, and gives a comprehensive summary of the phylogeny, anatomy, and adaptations of all marine mammal groups (modern & extinct), serving as an excellent introduction to graduate students interested in marine mammal paleontology. I got a copy of the second edition at the SVP benefit auction in Ottawa (2006) and read the entire front half - the second half, focusing more on soft tissues, diving behavior, and ecology and conservation is less relevant towards paleontologists. The third edition has nearly completely revamped the phylogeny and skeletal anatomy chapters, and many up-to-date references have been added. To my surprise, many citations of my own work were included - I'm young enough to still suffer from a bit of impostor syndrome, so seeing my papers referenced in a textbook is quite surreal. Several of my illustrations have been included as well, which I was quite pleased with! I highly recommend this for any "student" (in the broad sense) of marine mammal evolution.


The fossil record of pinnipeds was dominated by taxonomic confusion for nearly a century because most fossil pinnipeds - particularly the phocid seals - were named based on isolated postcranial bones and referred to the same species based on dubious criteria (see below, Koretsky et al.). Pinniped paleontology began in earnest in the North Sea, principally with true seals (Phocidae) and the overly confusing taxonomy of walruses (see my series on the walrus fossil record, hereXXX). Pliophoca etrusca, a Pliocene phocid from northern Italy, is a notable exception as the holotype is a partial skull with associated cervical, thoracic, lumbar, sacral, and even caudal vertebrae, as well as forelimb and hindlimb elements. However, it was originally reported in the 1940's and has been needing a modern "treatment" - which is exactly what this study provides. Pliophoca is very similar to extant Monachus (Hawaiian, Mediterranean, and the extinct Caribbean monk seals) but differs in several cranial, dental, and hindlimb features - such as having narrow, compressed incisors. This study reports many new specimens of Pliophoca cf. P. etrusca including mandibles, isolated teeth, and ankle bones, from Italy, France, and Spain. Previously, other material from the early Pliocene Yorktown Formation of North Carolina was referred to Pliophoca etrusca, but the authors rightly point out that no comparisons were made and no anatomical features linking the two were identified. Cladistic analysis fails to corroborate identification of this east coast USA material, which appears to represent an unnamed monachine instead; true Pliophoca is closely related to Monachus, sharing common ancestry and indicating origin of the Pliophoca-Monachus clade in the Mediterranean during the Plio-Pleistocene. This was followed by dispersal to the Caribbean, and later to the tropical Pacific.

The Pisco Formation is a spectacular Miocene shallow marine deposit in coastal Peru with excellent exposures of diatomite, sandstone, and mudrocks with skeletons of marine vertebrates littering the desert. Preservation is generally quite comparable to that of the Monterey Formation in central California in terms of preservation (i.e. abundant well-preserved skeletons with high rates of articulation), one of the only other stratigraphic units in the world where fossilized baleen has been reported - yet central California is not desert and is instead well-vegetated, relegating most exposures to difficult to access coastal outcrops. These two studies report highly detailed maps showing the occurrence of vertebrate fossils at two of the more important localities in the Pisco Formation: Cerro Los Quesos (cheese hills) and Cerro Colorado (red hills). The study at Cerro Los Quesos indicated that an earlier study conducted by young earth creationists failed to identify many smaller non-baleen whale fossils (small odontocetes, pinnipeds, birds, bony fish, sharks) calling into question the rigor of their field methods. At Cerro Los Quesos, the sheer majority of marine vertebrates are present within a 160 m thick section of the Pisco exposed at the tops of these hills. Again, at Cerro Colorado, marine vertebrates are concentrated into a narrow stratigraphic band (nearly 90% of all marine vertebrates from Cerro Colorado were found in a 35 m section near the base of their column). Both of these carefully executed studies fill in a desperately needed baseline for basic data in one of the world's premier marine vertebrate lagerstätten which, until the past year or so, was being supplied entirely by studies published by young earth creationists.

In the late 19th century a local found a large cetacean skull eroding from a cliff of the late Miocene Monterey Formation near Santa Barbara in southern California. It slowly eroded out over 30 years before being excavated in 1909. It was named Ontocetus oxymycterus by Remington Kellogg in 1925, who recognized it to be a very small part of a very large sperm whale. The tip of the snout is preserved with about ten upper alveoli, and the anterior tips of both mandibles with several poorly preserved teeth. The type species of Ontocetus, Ontocetus emmonsi, is not a sperm whale but is in fact a walrus and a senior synonym of Alachtherium and Prorosmarus. Because of this, O. oxymycterus needed a new genus name, and these authors provide a much needed redescription of this fragmentary but fascinating whale and rename it Albicetus oxymycterus. The teeth of Albicetus are enormous – 8 cm in diameter, roughly ¾ the size of the giant sperm whale Livyatan melvillei from Peru. I always assumed that the snout of this gigantic whale was incomplete, but these authors interpret the rostrum to be more than ¾ complete, and when plugged into body size calculations for odontocetes, a body size of 6 meters (~20 feet), small in comparison to the 10-14 meter length estimate of Livyatan despite the rather large teeth. The teeth of Albicetus, like Livyatan and other members of the “Scaldicetus” tooth grade, has enamel caps on the teeth. Given its size and robust enamel-capped teeth, Albicetus is probably another large macroraptorial apex predator like Livyatan – though in my opinion perhaps somewhat underestimated in terms of body size. 

The fossil record of eared seals (Otariidae) is limited and mostly consists of a few primitive fur seal-like species from the latest Miocene and Pliocene of the Pacific basin in comparison to their higher modern diversity and wider geographic range. The oldest known fur seals, Pithanotaria and Thalassoleon, are only known from the late Miocene of the North Pacific (California, Japan) and are no older than about 10 Ma in age; otariids evolved from an enaliarctine ancestor similar to Pteronarctos, but the youngest enaliarctines are much older - early middle Miocene, about 17-16 Ma. This ~7 million year gap in the mid Miocene begs the question "where the heck did otariids come from?" and "where were they in the middle Miocene?" Morgan and I published this paper after I discovered a partial mandible from the early middle Miocene Topanga Formation (~17.5-15 Ma) hiding in collections at the Cooper Center in Orange County. This specimen was misidentified as the small walrus Neotherium, but differed in having greatly simplified teeth and being much smaller - it was almost on its way towards being a typical late Miocene otariid, but retained an extra cusp on the lower molar lost in all other otariids (the metaconid) as well as primitively retaining a second lower molar. We named this new transitional pinniped Eotaria crypta, referencing its early age, and also its rarity - no specimens of true otariids have yet been reported from the ridiculously oversampled Sharktooth Hill bonebed, nor from well-sampled middle Miocene rocks in Japan. The earliest otariids may have been primarily pelagic, rarely straying into shallow shelf waters - a hypothesis originally proposed by our Japanese colleague Naoki Kohno.


The “river” dolphin Parapontoporia is widely known from the late Miocene and Pliocene of California, known from three species, and found exclusively in marine rocks. Its relationships have been elusive, and when first described originally thought to be similar to the La Plata River dolphin Pontoporia and placed in the Pontoporiidae – though similarities with the earbones of the now-extinct Yangtze River dolphin Lipotes were noted by the author. More recent studies utilizing cladistic methods have consistently identified Parapontoporia as a lipotid dolphin with no close relation to the pontoporiids (though the family is known from Pliocene rocks of the Atlantic states, and widely within South American fossil assemblages). Lipotes was completely riverine, which begs the question: why, when, and where did the lipotids become adapted to freshwater? In 2011 I found a single earbone, originally misidentified as a delphinid dolphin, hiding in fossil collections at UCMP in Berkeley. This earbone – the petrosal – is very distinctive, and closely matched those of Parapontoporia. However, this earbone was recovered from the non-marine Tulare Formation, which is late Pliocene in age, from the Kettleman Hills in the San Joaquin Valley of California. The valley was an ocean basin with a narrow connection until about 2 Ma when uplift of the Sierra Nevada introduced more and more sediment into the basin, along with uplift of the coast ranges which closed off the connection to the sea. At the time this individual of Parapontoporia died, the inland sea had transformed into a large lake or body of brackish water fed by rivers. This discovery not only indicates that palaeontologists in California should more closely search Pliocene terrestrial deposits for marine mammal remains, but that freshwater living may have characterized Parapontoporia in addition to Lipotes, heralding modern behaviour and distribution as far back as the late Pliocene.

[This is the third publication of my Ph.D. thesis - and one of the biggest chapters.] The history of study of eomysticetids is a bit convoluted - first formally recognized in 2002 with the publication of Eomysticetus, the most primitive toothless baleen whale. Eomysticetus was reported from the Oligocene of South Carolina, right here in the Charleston area - it has a mix of archaeocete-like features (e.g. tiny braincase, primitive earbones, large sagittal and nuchal crests, anteriorly placed blowhole, large fan-shaped coronoid process, "pan bone" on the mandible) and many derived features unique to modern baleen whales (rostral kinesis, flattened palate, toothlessness, beam-like mandibles without a fused symphysis). As it turns out, the story of eomysticetids really begins 70 years earlier with the discovery of an unassuming mysticete braincase in the Milton Lime Quarry in south Otago, about a half hour's drive from Otago Campus. It was named Mauicetus parki by Prof. Benham in the late 1930s; 20 years later, Prof. B.J. Marples discovered several mysticete skeletons in the Kokoamu Greensand and Otekaike Limestone of north Otago - and named all these as species of Mauicetus. One of these, Mauicetus waitakiensis, was reidentified as an eomysticetid and placed in the new genus Tohoraata last year (Boessenecker and Fordyce 2014). The most complete of these, Mauicetus lophocephalus, had an Eomysticetus-like skull; more recently, additional preparation and new specimens of Mauicetus parki shows that it is actually in the stem Balaenopteroidea - a "Kelloggithere", or cetothere sensu lato - poorly known baleen whales that are structurally similar to Parietobalaena. Because Mauicetus parki (the type species) and M. lophocephalus belonged to different families altogether, a new genus name was needed for the latter - but, always a complication: sometime in the 1960s an enterprising mover involved in moving the Otago Zoology dept. collections threw away the holotype skull of M. lophocephalus in the garbage, and is now likely in a landfill within 20 km of campus somewhere. R.E. Fordyce started collecting eomysticetid specimens from the Kokoamu Greensand and Otekaike Limestone in the early 1990s, and found at least one specimen (OU 22235) that is congeneric but with some tympanoperiotic differences (the tympanic bullae and postcrania of the type specimen were indeed spared by the ever-so-talented university movers) and another specimen (OU 22081) that was morphologically inseparable from the remaining elements of the type specimen. So, we named the first specimen as the holotype of Tokarahia kauaeroa, a beautiful Eomysticetus-like skull with a significant amount of postcrania, identified the second skull as a referred specimen of Mauicetus lophocephalus, and referred lophocephalus to Tokarahia, recombining it as Tokarahia lophocephalus. Tokarahia is structurally similar in terms of skull morphology to Eomysticetus but principally differs in having a longer occipital shield and more derived postcrania; Tokarahia kauaeroa has a beautiful mix of basilosaurid-like and modern mysticete-like features in the well-preserved tympanoperiotics and postcranial skeleton, and is a spectacular example of a transitional fossil. Most significantly, a single tooth with a flattened was found near the maxilla of the referred T. lophocephalus, matching the size and shape of maxillary tooth alveoli in other eomysticetids - suggesting that eomysticetids like Tokarahia retained a vestigial dentition, now representing an additional intermediate stage between the tooth/baleen bearing aetiocetids and the completely toothless crown Mysticeti.

At the time this study was published, eomysticetids from New Zealand (4 species: Tohoraata, Tokarahia), Japan (Yamatocetus canaliculatus), and South Carolina (3 species: Micromysticetus rothauseni, Eomysticetus spp.) were known, probably indicating a worldwide distribution with at least some local diversity (3 species from the lower Duntroonian stage of New Zealand, for example). Most of these are singletons – known by a single specimen (Micromysticetus rothauseni is a notable exception), meaning that patterns of adult variation and ontogenetic variation are unknown. This largely characterizes the cetacean fossil record in general, and particularly for Oligocene cetaceans. This study reports another beautifully preserved eomysticetid, Waharoa ruwhenua, represented by three partial skeletons including well preserved skulls, mandibles, and tympanoperiotics. The genus name Waharoa means “long mouth” in Māori, whereas the species name ruwhenua means “shaking land”, referring to the English name of the type locality - “The Earthquakes”. The adult has a very long surfboard-shaped palate, which is proportionally shorter in the two juvenile specimens, indicating that it nearly doubled in its length (relative to the width of the skull). Similarly, the mandible elongates as well from juvenile to adult. The tympanoperiotics are nearly adult-like in the juveniles, but the tympanic bulla appears to grow somewhat in size – a novel discovery amongst cetaceans, modern species of which all seem to be born with fully adult size earbones. These changes and  others reported in this study are among the first ontogenetic changes reported for any Oligocene cetacean. Additional features – and the ontogenetic trend of palate lengthening – permit inference of the feeding behavior of Waharoa. The anterior ¼ of the palate is barren and lacks palatal vascularization, suggesting that baleen was only present along the posterior ¾; the lengthening of the palate (more extremely than in any extant mysticete) suggests that extreme length is selected for in eomysticetids, and a feeding adaptation – rather than widening of the palate as in rorquals (minke, blue, humpback whales). Furthermore, the jaw joint appears to have been synovial and differs from the robust jaw joint of modern rorquals which permits extreme opening of the mouth, rotation of the jaws, and dislocation of the jaw joint – all this points towards Waharoa not being able to lunge feed like modern rorquals. Instead, the lengthening of the palate could be analogous to rostral arching of right whales as a strategy towards maximizing the cross-sectional area of the filter feeding apparatus – an adaptation for skim feeding. On our cladogram, right whales are the next diverging lineage after eomysticetids – potentially indicating that skim feeding is the primitive mode of feeding for baleen whales.

 Fossils of oceanic dolphins (Delphinidae) are not exactly common in late Neogene rocks of the eastern North Pacific. This family is the most diverse and widely distributed modern group of cetaceans and can be found in every ocean basin. Delphinids are widely known from north Atlantic and Mediterranean Plio-Pleistocene fossils, but are virtually unknown from the late Miocene except for a few fragmentary specimens from Japan (Eodelphinus) and California (unpublished). Given this background, I'm very interested whenever evidence turns up in California of fossil delphinids, since they are quite rare in Pacific margin sediments. In 2009, photos of a rather enormous skull which I at first thought was a monstrous beluga turned up in my email inbox; within a few days I had the collector on the line and he agreed to donate the specimen to UC Berkeley. Sometime later that summer I looked at a private collection and noticed two rather large pilot-whale like earbones, which the collector agreed to donate to UCMP as well. Most odontocetes from the late Miocene and Pliocene of California are small - porpoises, the "river dolphin" Parapontoporia, small delphinid dolphins, with only occasional evidence of early belugas and sperm whales. This rarity of large odontocetes, especially in the Purisima Formation of California - makes me quite interested whenever I stumble across any fossil evidence. After some preparation with an airscribe at Museum of the Rockies in Bozeman, Montana (where I was a student at the time), I found that the skull was actually rather similar to pilot whales (Globicephala) and false killer whales (Pseudorca) rather than extinct belugas (Denebola). The skull was found as float, but is almost certainly younger than 5.3 Ma and older than 2.47 Ma based on associated matrix and its likely stratigraphic position; the earbones (petrosals) were found in situ within a bonebed dated to 3.5-2.5 Ma. I initially thought the petrosals were from the same species as the skull, but a linear regression of petrosal size and skull size amongst delphinoids indicates that the petrosals are too small to belong to the same species as the skull - potentially indicating that two species of globicephalines inhabited the California coastline during the Pliocene. These new fossils, in concert with published fossils, indicates that globicephaline dolphins were already widely spread around the world by the early Pliocene.

Desmostylians are a bizarre but progressively more publicly beloved group of extinct marine mammals with hypothesized affinities to sea cows and proboscideans. Vaguely hippo-like with a conveyor-belt like tooth replacement seen in elephants and sea cows, and hippo-like tusks, desmostylians are inferred to have fed on sea grasses and kelp. The clade was never diverse, but new specimens are always being found and new material from the lower Miocene Unalaska Formation of Unalaska Island in the Aleutian island archipelago of Alaska appear to represent a new species. The new material includes a rather gigantic mandible similar to Desmostylus and Cornwallius, and several more fragmentary specimens. Based upon some minor differences with Cornwallius and Desmostylus, the new genus and species Ounalashkastylus tomidai was named. Another gigantic mandible known only as the Sanjussen specimen from Hokkaido is reidentified as a western Pacific occurrence of Vanderhoofius, formerly reported from the middle Miocene of California. The separation of Vanderhoofius from Desmostylus has been questioned before, but a distinguishing feature suggested by these authors is the loss of lower incisors during postnatal ontogeny; indeed, the upper incisors (but not lower) are lost by Desmostylus and Cornwallius during postnatal ontogeny. These desmostylines also bear large bony prominences on the medial side of the mandible but do not house unerupted teeth, and their function has remained unclear; this study suggests that either 1) the dense bone serves as ballast to keep the head negatively buoyant during feeding or 2) the bony prominences help buttress the mandible as the animal “clenched its teeth” together during suction feeding.  

The diet of pinnipeds is well-established for modern species, but it's difficult to determine the diet of fossil pinnipeds. For modern species we can go watch them eat – or we can cut them open when they die and look at what's in their stomach. For example, before we observed leopard seals filter feeding like crabeater seals, dead leopard seals with bellies full of krill had been found. Diet in fossil pinnipeds is difficult because we certainly cannot do the former, and the latter is rare – only one fossil pinniped with preserved gut contents has been recovered, the phocid seal Kawas from Patagonia. Diet, or at least feeding behavior, can be inferred some cases from feeding adaptations. Many pinnipeds feed in a similar fashion on fish (sea lions), whereas some others are primarily suction feeders that don't really use their teeth (walruses). Can diet be inferred from feeding morphology in extinct pinnipeds? This new study attempts to answer this question by using principal components analysis (PCA), hierarchical cluster analysis (HCA), and discriminant function analysis (DFA) to examine trends in tooth spacing and crown size, and diet. The DFA only reported a weak relationship with diet, and a stronger correlation between tooth spacing/crown size and feeding behavior (e.g. prey capture strategy. Tooth size and spacing were most strongly correlated with how important teeth were in prey capture, with narrowly spaced large teeth present in “biting” pinnipeds, and smaller, spaced out teeth present in “sucking” pinnipeds. Smaller tooth spacing and larger crowns also characterized pinnipeds that rip prey into pieces or filter feed (e.g. leopard, crabeater seals). This study applied these features to the extinct pinnipeds Desmatophoca and Enaliarctos, and recovered both as being similar to modern otariids – generalist feeders like sea lions and fur seals.  

Pinnipeds – seals, sea lions, and walruses – are a group of mammalian carnivores that evolved from dog or bear-like ancestors (or possibly otter-like – e.g. Puijila darwini). Modern pinnipeds are all either suction feeders or pierce feeders – teeth are used only to capture prey, but prey is swallowed whole instead of being chewed (masticated). Enaliarctine pinnipeds – the earliest known seals – have carnassials like terrestrial “fissiped” carnivores. Modern carnivores chew their food, and the carnassials shear through bone and flesh during mastication. Did enaliarctines chew their food like their terrestrial ancestors? And if so, did they have to leave the water to masticate after prey capture? To address these questions, data similar to that collected for modern and fossil pinnipeds as in Churchill and Clementz (2015: see above) was analyzed using principal components analysis (PCA) and phylogenetic independent contrasts (PIC) to see where Enaliarctos plotted within 2 dimensional "morphospace". Enaliarctines occupied an intermediate morphospace between terrestrial carnivores and pinnipeds, retaining close tooth spacing of "fissipeds", but with reduced heterodonty of pinnipeds. PCA indicated that Enaliarctos grouped with other pinnipeds as a pierce feeder - and that it likely did not masticate; this indicates that pierce feeding likely arose as a common feeding behavior of pinnipeds early during their evolution. Lastly, this suggests that Enaliarctos did not need to return to land after catching fish in order to feed, as suggested by some earlier studies. This article also gives a fantastic review of dental evolution of pinnipeds, supplementing earlier discussions by Boessenecker (2011) and Boessenecker and Churchill (2013). There's a common theme here: Morgan and I really like seal teeth!

On rare occasions vertebrate skeletons will get preserved with the remnants of their last meal. The identity of the gut contents associated with the Triassic dinosaur Coelophysis has been debated to death: are smallish bones found within the ribcage of an adult Coelophysis skeleton bones of a different species, or bones of a juvenile? Because the latter would make Coelophysis a cannibal. Fossilized examples of gut contents are rare, but provide pretty powerful data on trophic relationships - in other words, "who ate who". Examples of fossil marine reptiles with gut contents abound - from the Cretaceous of Kansas alone there is evidence of virtually every imaginable trophic relationship amongst large marine vertebrates. However, until this year there was very little evidence of gut contents for marine mammals. Two basilosaurids have been discovered with gut contents (Dorudon - published; Basilosaurus, not published) and two pinnipeds (phocid seal, New Zealand - private collection, not published; Kawas, published). Within Neoceti (baleen whales and toothed whales) there were no known examples. This new study reports fossilized gut contents of a late Miocene cetotheriid baleen whale from the Pisco Formation of Peru consisting of a mass of fractured fish bones tucked between the ribs of a partial baleen whale skeleton. The entire skeleton was not excavated by the authors, but the mass of fish bones was documented in situ and removed. The fish material consists of a single skeleton of a sardine, Sardinops; while the bones are fractured and disarticulated, they show no evidence of partial acid digestion. The authors interpreted the fish remains as being within the forestomach at death (whales are artiodactyls and thus have chambered stomachs). The authors interpret this as indicating that cetotheriid mysticetes were adapted to feed upon fish as opposed to soft bodied planktonic crustaceans (e.g. krill). A few minor problems exist with this study - such as the fact that the gut contents consist of a single individual - which is not a problem for the gut contents of a large macrophagous predator like a mosasaur, but for a filter feeder that would consume thousands of fish this size a day, it makes you wonder if it's an example of accidental ingestion. Another issue is that the taphonomically informative skeleton was left in the field (I suspect owing to a storage problem at the host museum). Regardless, it's a solid advance and I was pleased to see it published. Truth be told, I wondered when fossil odontocetes and mysticetes would be found with gut contents - and I always assumed they would be discovered in the Pisco Formation of all places. It's nice when predictions are verified!

Modern true porpoises (Phocoenidae) are amongst the smallest of all cetaceans, and few surpass 2.5 meters in length; they share a common ancestry with oceanic dolphins (Delphinidae) and white whales (Monodontidae) sometime during the middle or late Miocene, perhaps arising from the "kentriodontid" dolphins. The sheer majority of fossil porpoises are from the north Pacific with a few important specimens from the west coast of South America, a possible periotic from New Zealand, and an extinct genus Septemtriocetus from Belgium. In contrast, phocoenids are currently nearly worldwide in distribution (within subtropical/temperate waters, anyway). This study reports a second porpoise from the North Sea, Brabocetus gigaseorum, based on a partial braincase from the early Pliocene Kattendijk Formation of Belgium. In many regards this genus is Phocoena-like (harbor porpoise) with a similar facial region but differs by possessing some archaic features. At first glance, I would assume that this would be one of the closest morphological matches to modern phocoenids, which typically form a crown clade without any extinct genera in cladistic analyses of porpoises, with all extinct genera of porpoises falling outside this group. However, their analysis shows Brabocetus forming a clade with Septemtriocetus, Haborophocoena, Salumiphocoena, Archaeophocoena, Miophocaena, and strangely, Semirostrum. I'm skeptical of the placement of Brabocetus, and I strongly suspect that the Phocoenidae is taxonomically oversplit - but more fossils are the only way to cure this issue and this paper is a fine contribution to porpoise evolution. Because Brabocetus and Septemtriocetus are in the eastern North Atlantic, they suggest an early Pliocene dispersal of phocoenids through the Arctic shortly after the opening of the Bering Strait - followed by a second dispersal through the arctic during the middle or late Pleistocene by extant harbor porpoise (Phocoena phocoena).

The giant sea cow Hydrodamalis gigas was discovered by the shipwrecked crew of the Svyatoy Pyotr in the Komandorsky Islands in 1741 and named the Steller's sea cow after the Russian expedition's German naturalist Georg Wilhelm Steller. Within 30 years this giant kelp-feeding sirenian was extinct; for over two centuries it was assumed that Hydrodamalis was hunted to extinction because of how easy it was to kill. Indeed, stories about this source of food circulated amongst fur traders in the Kamtchatka region, and it has always been suspected that subsistence by fur traders drove the last population of Hydrodamalis to extinction. Another hypothesis that has gained traction in recent years, but has been notoriously difficult to actually test - is the possible influence of sea otter hunting rather than direct hunting of Hydrodamalis. After all, Hydrodamalis lived pretty much from Japan to the Aleutians and down to Baja California during the late Pliocene, long before humans ever made it to the Pacific coasts; whatever snuffed out the last remaining populations of Hydrodamalis in the subarctic was perhaps a long time coming. The idea is simple and elegant: sea otters tend to keep sea urchin populations down, and in areas where sea otters have been removed from the environment by overzealous hunting, sea urchins completely consume and destroy kelp forests (within 5-8 years of sea otter extirpation). The diet of Steller's sea cow was entirely based on kelp - and the crew of Bering's expedition noted abundant sea otters in the Komandorsky Islands. Because of this relationship, sea otters are a keystone species and help maintain kelp forests. Whereas sea cows were demonstrably extinct by 1768 at Bering Island, sea otters had been extirpated in the Komandorsky Islands by 1753; in Alaska, sea urchin populations did not really "explode" until a few years after sea otters began to decline, followed by kelp forest collapse. This lag matches rather well with the earlier reported extirpation date of sea otters (1753) and the sea cows (1768). [Note that the Bering expedition discovered the sea otter as well, and within a year or so of being discovered the word got out about their luxurious fur and the sea otter fur trade began; sea otters were hunted mercilessly from west to east, with the Russians pushing the fur trade into Northern California by the early 19th century.] The authors go one step further and used population modeling and simulation of starvation to show that sea otter hunting alone, even without any direct hunting of sea cows, would have driven extinction of the doomed giant sirenian by itself.

This paper is a review of archaeocete evolution, and since it is a review paper, will only get brief treatment here. The review is intended to provide a comprehensive summary of trends in archaeocete evolution - indeed, nearly 2/3 of the paper consists of a family-by-family discussion of what we know about each archaeocete "family". While the reference list is quite good in terms of inclusion regarding papers published prior to 2005, many more recent advances such as the discovery of good postcrania for North American protocetids (Natchitochia, Georgiacetus), protocetids with excellent skeletons and fetuses (Maiacetus), and also fails to include any citations of recent excellent work produced by Julia Fahlke and colleagues. The paper casts doubt on the hippo-whale link, consistently claiming that the cetacean-mesonychid hypothesis is always identified when molecular data is excluded - which may have a grain of truth, but the most comprehensive analysis to date by Michelle Spaulding et al. (2009: PLoS One) showed that tree topology of the artiodactyls (and thus cetaceans and mesonychids) is very sensitive towards which taxa are included/excluded from the matrix, which seems to be a bigger problem than just molecules v. morphology/combined analyses. Ultimately, this paper doesn't really present any new ideas, and since we've already had on average one review paper on archaeocete evolution ever two years, I'm not sure why we needed another - particularly considering that many important studies of archaeocete evolution have been omitted in this study. The authors could have just consulted my 2012, 2013, and 2014 marine mammal paleontology summary posts!

Marine vertebrate paleontologists studying fossils from the west coast of North America are well-acquainted with concretions - extremely hard carbonate (and occasionally phosphatic) nodules that form around marine vertebrate remains. Concretions are much harder than surrounding rock, and will often erode out of cliffs but keep the entombed fossil in good condition as it slowly worries away by wave action. Concretions are an absolute pain to work with, as they often require mechanical preparation with pneumatic tools or chemical preparation with acids to remove the bones. Fossil marine vertebrates in the Pisco Formation of Peru are often entombed within dolomite concretions, and may be linked with the excellent preservation of fossils. This new study is yet another excellent contribution towards the geological context of spectacular marine vertebrate assemblages from the Pisco Formation, and reports new data from the field and petrographic results to investigate the formative processes involved in Pisco concretions. These authors note that a higher proportion of mysticete (baleen whale) fossils are associated with concretions, suggesting that they are more likely to form with larger carcasses. Specimens in concretions are also more likely to be complete and articulated. Based on the distribution of dolomitic matrix, soft tissues must have already been decayed prior to formation of dolomitic cement, and bones must have been at least partially buried. In some cases, dolomitic matrix forms only within the bones - an ideal situation, as it  has prevented burial compaction and deformation of the bones. Dolomite is commonly assumed to be a diagenetic or metamorphic "sequel" to original limestone - but in these cases, it appears that dolomite was directly precipitated without a limestone precursor (a similar process affects Purisima Formation vertebrate-bearing concretions at Point Reyes National Sea Shore). This study proposes that dolomite was precipitated as a response to the decay of organic matter of the whale after skeletonization and burial, thereby forming nodules and infilling bones with cement that contribute to their preservation and recovery. An earlier study by young earth creationists claims that excellent vertebrate preservation in the Pisco Formation is caused by extremely fast sedimentation (these rates, by the way, they use in the non-scientific literature and extrapolate to the entire Pisco basin, claiming that the entirety of the strata would have been deposited in 20,000 years rather than 20 million, supposedly meaning that scientific dating methods do not work). Instead of requiring bizarre claims about ultra-fast (shall we say, biblical?) sedimentation rates, this new study finds a far more logical solution: that the organic matter whales possess at the time of death and burial contributes to dolomitic cementation and ultimately their preservation.

In 2013 Mark Uhen published a review of basilosaurids from North America and noted that the classic species Zygorhiza kochii, the smallest basilosaurid from the Eocene of the gulf coast, is based on a poorly preserved braincase that does not preserve any autapomorphies and is therefore non diagnostic, thereby making Zygorhiza kochii a non-diagnosable taxon. To remedy this Uhen submitted a proposal to the ICZN requesting that well known specimen USNM 11962 (the de facto reference specimen for the species since Kellogg described it in the 1930s) be designated as a neotype. Gingerich (alluded to above) published this comment on Uhen's proposal effectively pointing out that such actions are unwarranted under current ICZN rules for type specimens. Type specimens do not necessarily need to be "good" or diagnostic (they just ought to be). Gingerich (and see Gingerich, above) insists that the Jackson Group cetacean assemblage includes only three basilosaurids: a small species (Zygorhiza), a medium sized species (Pontogoneus brachyspondylus, aka Cynthiacetus maxwelli), and a large species (Basilosaurus cetoides). Under this paradigm, all basilosaurid material smaller than Pontogoneus/Cynthiacetus belongs to Zygorhiza anyway, and designating a neotype wouldn't really solve any pressing issue. This debate highlights an interesting philosophical rift within paleontology: what good are shitty type specimens? I see pros and cons on each side, but at the end of the day, identification of new material should be based on more than just size alone, meaning that having anatomically informative type specimens is important towards reliably making comparisons and identifications of new material in a semi-repeatable, non-willy nilly approach.

The toothy basilosaurids are easily some of the most iconic of all cetaceans owing to their fearsome jaws and large size, and are also the largest and most completely known of all the archaeocetes. Many are quite large (Basilosaurus, Basilotritus) but others are smaller - including Dorudon and Zygorhiza. The former is easily the most well-known archaeocete (South Carolina, Egypt) and the latter is well-known from the southeastern USA - but this paper was more or less the first treatment of Zygorhiza since Remington Kellogg's seminal masterpiece "Review of the Archaeoceti" published 80 years ago, a fact which Gingerich laments. There are many new specimens of Zygorhiza now, but for some strange reason very few of them have been published upon (the same can be said for Basilosaurus cetoides, in my opinion). A "new" specimen (collected in a large plaster jacket in the 1960s, and shopped around on loan for forty years until Gingerich took it to Michigan to get prepped out) includes part of a skull and much of the vertebral column, allowing reevaluation of vertebral numbers and morphology. Further, the skull was scanned to produce a digital endocast, which was compared with a digital endocast for a separate specimen published a decade ago by Lori Marino and others. Relative brain size is on the small size for a terrestrial mammal of comparable body size, but intermediate between mysticetes and odontocetes. Lastly, Gingerich discusses the archaeocete fauna of the Eocene of the Gulf Coast, remarking that there is always one small basilosaurid (Zygorhiza kochii), one large basilosaurid (Basilosaurus cetoides), and a medium size basilosaurid. Pontogeneus priscus was originally based on an isolated, medium sized cervical vertebra, in the 19th century when standards for paleocetological holotypes were not yet "evolved". This taxon was later declared a nomen nudum by Mark Uhen, who named the new genus and species Cynthiacetus maxwelli based on a partial skeleton including a nearly complete (but incompletely figured) skull. Gingerich interpreted Cynthiacetus as a junior synonym of Pontogeneus, and discussed what we can and should do with ICZN rules for types. In other words: a type need not survive to the present day (see below) or even be diagnostic, so we can get away with using shitty Georgian or Victorian era types. My opinion: if there's no preserved autapomorphic features (e.g. the fossil is a cetacean vertebra), it should be declared a nomen dubium as it is impossible to unambiguously diagnose the taxon. A name, then, is only as good as its type. We'll see how this one pans out.

A lot of ideas are floating around in this paper, which reports a new archaeocete fauna from the middle Eocene Aridal Formation of Morocco. Six archaeocetes are present, including three protocetids (protocetids are the geochronologically latest archaeocetes which retain hindlimbs - Maiacetus, Rodhocetus, Georgiacetus). The protocetids include unidentified small and medium sized taxa based on fragmentary postcranial bones, and teeth and postcrania of the large protocetid Pappocetus lugardi. Pappocetus is of particular note because it was originally reported from similarly aged deposits in Nigeria, and this record extends the range of this early whale across most of west Africa. The basilosaurids tell a much more interesting story, and include the small bodied new species Chrysocetus fouadassii, the somewhat larger new species Platyosphys aithai, and the large basilosaurid Eocetus schweinfurthi. The new species of Chrysocetus is notable since the type species, Chrysocetus healeyorum, was named from the Eocene of South Carolina, and is the smallest basilosaurid from the Atlantic coastal plain - and thought to be one of the earliest monophyodont cetaceans (e.g. only a single set of teeth). Most of the differences with C. healeyorum are expressed in the postcranial skeleton. I'll explain Eocetus next, since it's a bit easier: Eocetus schweinfurthi, formerly identified as a giant protocetid, is likely instead a basilosaurid according to this study. New specimens include teeth, vertebrae, and a tympanic bulla - and expand the range of Eocetus schweinfurthi from Egypt to Morocco. Lastly, a strange basilosaurid skull with abundant pachyosteosclerotic growth and large, inflated vertebrae with many small foramina and elongate transverse processes is named as Platyosphys aithai. It matches the vertebral anatomy of Platyosphys paulsonii, a poorly known archaeocete described in the late 19th century from the Eocene of Ukraine. The type of P. paulsonii is an isolated vertebra that is now lost. Platyosphys-like species have since been reported from the southeastern US (Eocetus wardii) and then Ukraine again, and named Basilotritus uheni (and the former reclassified as Basilotritus wardii). Gol'din and Zvonok (2013) declared Platyosphys to be a nomen dubium because the type specimen is lost; however, Gingerich points out that technically speaking, under the ICZN, a name is available so long as evidence towards the type actually existing has been published - so despite being lost, Brandt (1873) published beautiful illustrations of the type vertebra(e). Is the type specimen crappy? Yes, and it's possibly gone forever - but the vertebrae of Platyosphys paulsonii are very distinctive in comparison to the crappy type specimen of Pontogoneus brachyspondylus (see above) so I'm more inclined to agree with Gingerich on this case.

This new paper provides a long-needed redescription of the problematic "cetothere" Mesocetus argillarius, originally named by Flemming Roth in the late 1970s. This specimen is from the upper Miocene Gram Formation of Denmark, which historically has also yielded the Pelocetus-like baleen whale Uranocetus and poorly preserved pontoporiid dolphins. Differing from other Mesocetus (which similarly deserve modern re-treatment), the authors assign it to the new genus Tranatocetus. The skull is broadly similar with some poorly known "cetotheres" (occasionally regarded as true Cetotheriidae or "cetotheres" sensu lato) such as Mixocetus, "Aulocetus" latus, "Cetotherium" vandelli, and "Cetotherium" megalophysum,, but has a very primitive lower jaw with an enormous mandibular foramen. Cladistic analysis places these poorly known "cetotheres" including Tranatocetus as sister to the Balaenopteroidea (gray whales + rorquals), and the authors erect a new family, Tranatocetidae, based on derived features of the tympanic bulla and some other features of the braincase. "Cetotheres" in this study have been split into the Cetotheriidae, sister to the Neobalaenidae similar to the provocative hypothesis of Fordyce and Marx (2013), on the neobalaenid/cetotheriid "stem", within the Tranatocetidae, and on the tranatocetid + balaenopteroid "stem". Further testing of this interesting hypothesis of mysticete relationships will require redescription and reanalysis of poorly known "cetotheres" like "Cetotherium" megalophysum and "Cetotherium" vandelli.

Bone histology is a useful way to study how vertebrates grow. In the case of many terrestrial vertebrates, bone growth can even be studied at a level where annual growth lines may be counted like tree rings - many friends of mine who were in graduate school when Sarah and I were at Montana State involved this sort of study - Holly Woodward, Julie Reizner, Alida Bailleul, John Scannella, and even our friends Liz Freedman-Fowler and Laura Wilson-Brantley (taphonomists originally!) couldn't avoid histology. Marine vertebrates on the other hand do not preserve these sorts of growth lines quite so well (though counting growth lines in marine mammal teeth is a commonly used method for modern species). However, various marine tetrapods have adapted their terrestrial skeletons to problems of swimming and buoyancy in various ways, and many possess peculiar patterns of dense bone growth, hypothesized by some to be bone ballast. Histology is destructive, and this study sought to sample many postcranial bones of archaeocete whales including precious vestigial hindlimbs - so high resolution CT imaging was used instead to study changes in bone microstructure across the terrestrial-marine transition in early whales - remingtonocetids, protocetids, and basilosaurids. All exhibited bone mass increase in their ribs, but the vertebral column consists chiefly of spongy bone. The humerus of protocetids retains thick cortex (an adaptation for locomotion on land or paddling), but within basilosaurids the humerus becomes more strongly spongy and porous - indicating a transition from forelimb paddling to use of the foreflipper as a simple hydrofoil like modern cetaceans. On the contrary, the femur becomes very dense - even in the vestigial hindlimb of basilosaurids, which remains unexplained. The pattern of bone mass increase in Basilosaurus was originally interpreted to be an adaptation for control of "trim" - orientation of the vertebra column with respect to the horizontal plane (e.g. "pitch" in an aircraft) - sampling additional bones from the skeleton now rules out this hypothesis, but the authors indicate that no other existing explanation is sufficient. Regardless, patterns of bone microstructure are consistent with remingtonocetids and protocetids being shallow diving, semiaquatic swimmers with a limited capability of terrestrial locomotion like pinnipeds and sea otters, whereas basilosaurids are broadly comparable with modern cetaceans, consistent with interpretations of basilosaurids being the earliest oceangoing (pelagic) cetaceans. This study also showed that microstructure can change dramatically along the long axis of a rib or other bone, demonstrating the importance of taking numerous sections.

How did the skull of aquatic carnivores evolve after making the land to sea transition? Was it a passive process, or did pinnipeds undergo an adaptive radiation? This new study by Katrina Jones and others investigates this by using 3D morphometrics of modern and extinct terrestrial "fissiped" and pinniped carnivorans within a phylogenetic context. Several fossil pinnipeds were included such as Enaliarctos emlongi, Allodesmus, Pontolis, Piscophoca, and Acrophoca. Overall pinnipeds exhibit a greater variation in skull shape (disparity) than terrestrial carnivores. However, there is no increase in evolutionary rate at the base of the pinniped tree, indicating passive evolution of skull shape (e.g. "Brownian motion") and indicating that an adaptive radiation model does not fit very well. Within later groups of pinnipeds evolutionary rates sped up, perhaps associated with ecological specialization (e.g. walruses and suction feeding). In the context of the late Oligocene-early Miocene pinniped fossil record, this does make sense as nearly all pre-middle Miocene pinnipeds are enaliarctines that look fairly similar and share similar body sizes despite belonging to different lineages (Enaliarctos, Pteronarctos, Pinnarctidion, Prototaria, Proneotherium) - Desmatophoca being a notable exception.

This article is entirely in Japanese, and while it does have an English abstract and good photos, the fossil described is incomplete and the abstract short – so my summary will be as well! This specimen is from the middle Miocene and consists of a partial posterior part of a mandible. This specimen has a small, laterally projecting coronoid process, an enormous and anteriorly expanded mandibular foramen a tiny mandibular condyle, and a ventrally deflected angular process. These features are all consistent with this specimen belonging to a “Kelloggithere” - an informal name for “cetotheres” sensu lato like Parietobalaena, Pelocetus, and Diorocetus. This group is well-represented by a large collection of terribly understood fossils. Other specimens with a similar angular process have also been reported from Japan, but belong to an unknown mysticete – a similar mandible is present in Mauicetus parki from New Zealand.

This article is also entirely in Japanese with the exception of the abstract, so my summary is going to be brief! It does have an English abstract and good figures, so I'll communicate what I can. Japan has an excellent fossil record of cetaceans including my favorite group - baleen whales. However, many of these are not yet described, and many late Miocene and Pliocene mysticetes from Japan have been under study for years without any resulting publications, a frustrating situation for some. This new paper reports a balaenopterid whale, more commonly known as a rorqual (humpbacks, minke, fin, blue whales are all rorquals), from upper Miocene rocks of Miyako Island which is located pretty far south and relatively close to Taiwan. This specimen consists of a partial skull lacking a rostrum, is somewhat fractured, and has some adhering concretionary matrix. The specimen cannot be identified to any existing balaenopterid genus owing to its braincase morphology - and is too incomplete to designate as a type specimen, so the authors simply identify it to the family level. In my opinion, this whale is most similar to the archaic taxon Protororqualus cuvieri, but it's anybody's guess at this point; more preparation would be instructive, as the critical earbones are still in situ. A tantalizing find, and according to Felix Marx more research is being done on this specimen.

Many of the world's oldest well-preserved phocid seals (true, or earless seals) - and many of the earlier reports of fossil true seals in general - come from marine deposits of Paratethys, a former sea that occupied a foreland basin to the north of the Tethys sea (the Black, Caspian, and Aral seas are the remaining deep pockets of this former sea). Paratethyan deposits stretch from Austria to Kazakhstan; biostratigraphy is often rudimentary, but marine mammal fossils are plentiful. One of the earliest phocids known by a good skull was named Devinophoca claytoni in 2002; this early seal has features of both the Monachinae (southern seals) and the Phocinae (northern seals), potentially indicating ancestral relationships. This new article by Irina Koretsky and Sulman Rahmat names a second species of Devinophoca based on another well preserved skull, isolated teeth, and mandibles. The skull is similarly generalized, but has a monachine-like number of incisors; the new mandible differs from monachines and is very similar to the gracile mandible of the early phocine Leptophoca. Interestingly, Devinophoca inhabited subtropical waters with abundant corals, and was likely a shallow diver.

Unlike the above mentioned Devinophoca, the sheer majority of the phocid (true seal) fossil record is constituted by disarticulated, non-associated postcranial bones. This problem has plagued pinniped paleontology in general since J.P. Van Beneden began studying Mio-Pliocene true seals from North Sea deposits in Belgium and the Netherlands, and named a bunch of problematic species & genera based on disparate material, many of which have been suspected to be chimaeric assemblages of postcrania. With the exception of European & Paratethyan taxa based on crania (Devinophoca, Praepusa, and Pliophoca - see Berta et al., above) and New World taxa based on associated skeletons (Monotherium wymani, Leptophoca, Piscophoca, Acrophoca, Hadrokirus, and now Australophoca - see below) it's unclear how many of these postcranial taxa are real. How much do pinniped postcrania vary within a population, or how similar are they between taxa? Sexual dimorphism is also a huge problem. These are valid but also very basic questions that have not really been addressed, yet the study of fossil phocids has been plagued by these problems for over a century while students of true seal anatomy continue to ignore them. This new study reports several isolated postcranial elements from the including two humeri and a sacrum - one complete female humerus, and a partial male humerus. These remains do reflect an extraordinarily tiny seal - smaller even than the newly described Australophoca (see Valenzuela-Toro et al., below). The authors make favorable comparison to postcrania of the Miocene Paratethyan seal Praepusa, and base the new species Praepusa boeska on the complete female humerus and refer the other bones to it. Dating of the locality is poor - late Miocene to "mid" Pliocene, 11.5-3.5 Ma. The authors discuss the fossil record of Praepusa, and point out that the earliest fossils (P. vindobonensis) originate from the middle to late Miocene (16.5-11.2 Ma) of Kazakhstan and Austria, the somewhat younger species P. pannonica is from the late Miocene (12.3-11.2 Ma) of Moldova and Hungary, P. magyaricus is similarly from the late Miocene (13.6-12.3 Ma) of Austria, and the new North Sea species, P. boeska, is from the late Miocene-Pliocene of the eastern North Atlantic (11.6-3.2 Ma). This does paint a rather interesting picture of seal dispersal out of the Paratethys westward into the north Atlantic. Future discoveries of cranial material and associated skeletons are needed to assess whether or not Praepusa is monophyletic.

The early Miocene is an intimidating time for students of toothed whale evolution – quite a bit is going on, and there are a zillion different types of long-snouted dolphins living alongside the early ancestors of modern groups (some of the earliest sperm whales, the earliest possible delphinoids, purported early beaked whales). Some of these speciose families of longirostrine dolphins either originating or diversifying during the early Miocene include the Eurhinodelphinidae, Squalodelphinidae, Platanistidae, Allodelphinidae, Eoplatanistidae, and the “Dalpiazinidae”. Early Miocene odontocetes suffer from several problems – many are known from good skulls but not well-prepared or figured earbones; some taxa are almost certainly oversplit, and there is likely an overemphasis on the definition of family level taxa, some of which are likely para- or polyphyletic. Lastly, because of these issues, the phylogenetic relationships of these problematic dolphins are poorly known – and these reasons are why the early Miocene scares the shit out of me. This study daringly reports a long snouted dolphin, Chilcacetus cavirhinus from the lower Miocene Chilcatay Formation of Peru. This dolphin has a long snout and a homodont dentition – in other words, the teeth are all identical in shape and all are single rooted. Chilcacetus uniquely has a deep cavity between the nasal bones and the mesethmoid. It is similar to the giant dolphin Macrodelphinus kelloggi from the lower Miocene Jewett Sand of California, which has been classified in the past as a giant eurhinodelphinid. However, Chilcacetus – which has part of a mandible as opposed to the fragmentary Macrodelphinus type specimens – has an unfused mandibular symphysis, unlike all eurhinodelphinids. The taxonomically informative earbones were unfortunately lost between collection and publication, but somewhat detailed line drawings were prepared before they were lost. Other features preclude assignment to any other odontocete family. Cladistic analysis of Chilcacetus and other odontocetes actually supports a clade including Chilcacetus, Macrodelphinus, Argyrocetus from Argentina, and “Argyrocetus” (two species from California) – which could be named as a new family. The authors stop short of this, highlighting low statistical support for the grouping and lack of unambiguous synapomorphies. However, the grouping does consist of mostly eastern North Pacific and South American species; anatomical features supporting this clade are mostly primitive features that set them apart from later diverging odontocetes. As per usual, more fossils and more character evidence is needed to make sense of these challenging taxa.

Gut contents is widely reported for fossil skeletons of marine reptiles – skeletal remains of a larger animal's last meal. Mosasaurs, plesiosaurs, ichthyosaurs, sharks, and large fish are all known with well-preserved gut contents. For some reason, however, gut contents of marine mammals is much more rare. Ironically, little is known about the feeding behavior of modern marine mammals since it is difficult to directly observe them out at sea – so much of our knowledge of the diet of modern marine mammals is recorded from necropsies and gut contents. The difference is that in modern specimens, soft bodied prey can be observed – but in the rock record, only prey items with hard tissues become preserved, with some rare exceptions (e.g. cephalopod beaks, which are keratinous). Prior to this study the only marine mammal with gut contents was the Miocene seal Kawas from Argentina. I have wondered when the first fossil cetacean with gut contents would be discovered, and figured that if it were to be discovered anywhere it would be from the Pisco Formation of Peru. My gut instinct (no pun intended) was vindicated this year with the discovery of a Messapicetus specimen with abundant sardine skeletons preserved around it. However, it is unclear what exactly was collected as opposed to left in the field (see Collareta et al., above). Regardless of these issues, this study indicates that an early beaked whale (Ziphiidae) had a gut full of shallow water fish (Sardinops) when it died. Modern beaked whales are suction feeding, deep diving squid specialists. Assuming that this individual died after consuming a typical meal for its species, this fossil may indicate that Messapicetus was not a deep diving squid specialist, perhaps indicating that deep diving “teuthivory” is a more recent feature of beaked whales. Surely, the Pisco Formation will produce more treasures that will expand our knowledge of ancient food webs.

This new study is the product of Felix Marx's Ph.D. thesis at the University of Otago, and I was fortunate to see quite a bit of this long before it came out. The core of this study is a new phylogenetic analysis of baleen whales (Mysticeti) and is a successor to an earlier cladistic matrix produced during Marx' master's program at Bristol. This new phylogenetic analysis includes the most number of baleen whales (though with 3/4 the character evidence as the largest published analysis). These authors again recovered an aetiocetid-mammalodontid toothed mysticete clade, a monophyletic Cetotheriidae including the pygmy right whale, possible resolution amongst "kelloggitheres", and gray whales (Eschrichtiidae) deeply nested within the rorquals (Balaenopteridae). This new analysis is essentially a semi-comprehensive study of mysticete evolution, and also used the morphological dataset to study disparity (anatomical diversity) through time. Disparity peaked during the Oligocene and plateaued during the Neogene, whereas taxonomic diversity was highest during the middle and late Miocene and dropped off during the Plio-Pleistocene (perhaps an artifact of the rarity of published accounts of Pliocene mysticetes). Alternatively, evolutionary rates were highest during the Oligocene and flatlined thereafter - suggesting early settling into "modern" filter feeding niches. Mysticete diversity seems to drop as soon as modern gigantism appears, suggesting an influence of Plio-Pleistocene glacial influence on baleen whale evolution.

Marx, F.G., C.H. Tsai, and R.E. Fordyce. 2015. A new early Oligocene toothed ‘baleen’ whale (Mysticeti: Aetiocetidae) from western North America: one of the oldest and the smallest. Royal Society Open Science2:150476.

Most modern baleen whales are known best for their massive size – the smallest modern baleen whales – the pygmy right whale Caperea marginata and the dwarf form of the minke whale Balaenoptera acutorostrata – are still 5-6 meters long. Many of the earliest baleen whales which retained teeth – aka toothed mysticetes – were on average quite a bit smaller. One family in particular, the Aetiocetidae – perhaps the best understood of all toothed mysticetes next to the well-published mammalodontids – were comparable in size to modern small-bodied dolphins. Aetiocetus & Morawanacetus are in the size range of bottlenose dolphins, whereas Chonecetus is approximately the size of a harbour porpoise. Aetiocetids are hypothesized to be the earliest baleen-bearing cetaceans, with teeth and palatal vascularization reported in Aetiocetus weltoni. However, their earbones – an anatomically informative part of the skull – are very poorly known, and their teeth are virtually unknown aside from two species of Aetiocetus and a single pair of molars in Morawanacetus, making inferences about their diet and feeding adaptations difficult. A new specimen from the Makah Formation of the Olympic Peninsula in Washington is named as Fucaia buelli – the species name in honor of the talented and prolific illustrator Carl Buell (a well deserved honor, and I think this was a very nice touch by the authors), with the genus name referring to the type locality along the southern shore of the Strait of Juan de Fuca. This new aetiocetid preserves a partial skull, earbones, some teeth, many vertebrae, and a scapula – it is characterized by small size, large eye sockets (orbits), primitive archaeocete-like teeth, delicate hyoid bones. Cladistic analysis allies this species with Chonecetus goedertorum, which the authors reassign to the new genus as Fucaia goedertorum. Though most of the feeding apparatus is gone, the teeth are heterodont – caniniform anterior teeth and multicuspate, subtriangular cheek teeth are present, similar to archaeocetes – and they have wear facets, indicative of some degree of occlusion. Tooth morphology and wear suggests that Fucaia was a raptorial feeder, using its dentition to catch fish, contrasting with the comparably tiny teeth of Aetiocetus (Aetiocetus is approximately double the size of Fucaia, but has teeth that are much smaller) – the latter was likely an early filter feeder owing to its tiny teeth, rotate-able mandible, and possible presence of baleen. The authors suggest that aetiocetids (and mammalodontids) represent an intermediate stage of feeding adaptation involving raptorial suction feeding, which led to suction-based filter feeding, and eventually “pure” filter feeding.


Desmostylians are known only from the late Oligocene and Miocene of the North Pacific, ranging from Japan north to Alaska (see Chiba et al., above) and south to California and Baja California. Paleoparadoxia has of course now been split out into three genera, including Archaeoparadoxia and Neoparadoxia. Prior to this, Paleoparadoxia was the longest ranging desmostylian, ranging in age from about 23-9 Ma or so. Since being taxonomically split, the situation is a bit different - and this study seeks to refine the geochronologic range of true Paleoparadoxia, as the splitting by Barnes (2013) mostly affected fossils from the Miocene of California, but left the situation unclear regarding other paleoparadoxiines. This study reviews the fossil record of "true" Paleoparadoxia from Japan and reports a partial forelimb from the earliest Miocene. Several records of Paleoparadoxia from Japan are associated with cold temperate mollusk assemblages while others are associated with warm temperate mollusks - indicating flexibility towards water temperature (i.e. within the lineage). The new specimen reported from Hokkaido is the geochronologically oldest record of Paleoparadoxia, determined to be earliest Miocene (23.8-20.6 Ma) by the authors. Inuzuka previously hypothesized that Paleoparadoxia from the early Miocene of Japan may have evolved directly from Archaeoparadoxia weltoni. However, the discovery of a more derived true Paleoparadoxia from temporally equivalent rocks in Japan indicate that at least two paleoparadoxiines were living during the earliest Miocene, complicating the picture somewhat.

Modern river dolphins were formerly thought to constitute a single group, the Platanistoidea - until it became obvious in the 1980s from skeletal anatomy and fossils that each modern river dolphin (Inia, Lipotes, Platanista, Pontoporia) likely had separate origins, later confirmed by molecular work in the early 2000s. Two of these may in fact be somewhat closely related - the Amazon river dolphins, Inia, and the La Plata river dolphin, aka Franciscana, Pontoporia (which is actually mostly marine). The recently extinct Yangtze river dolphin Lipotes is now thought to be closely related to Parapontoporia from California (see Boessenecker and Poust, above), but has no other near relatives in the fossil record. The origins of the totally bizarre Ganges/Indus river dolphins Platanista have been more contentious, possibly involving the Squalodelphinidae, Squalodontidae, and even the Oligocene Waipatiidae. The evolution of the South American river dolphins is less controversial, and many fossils have been reasonably identified as extinct marine (and freshwater) of Inia (Ischyorhynchus, Saurocetes, Goniodelphis, Meherrinia) and Pontoporia (Brachydelphis, Pliopontos, Protophocoena, Auroracetus, Stenasodelphis). A new fossil discovered by J. Velez-Juarbe and others during the Panama Canal Project (PCP-PIRE) from the late Miocene of Panama includes a large, well-preserved skull and mandible with associated teeth, scapula, and carpal elements and is named in this study as Isthminia panamensis. It is large (nearly 3 meters in length based on skull size) with relatively robust teeth and an elongate, narrow rostrum like extant Inia. Unfortunately, anatomically informative tympanoperiotics are unknown. Because Isthminia was recovered from marine sediments, the authors interpret it to be a marine rather than a freshwater iniid. However, it is important to note that terrestrial animals frequently become entombed in marine sediments (likely carried out to sea during floods) and in general tetrapods are terrible paleodepth indicators. Regardless, the cladistic analysis and inferred environmental preference of modern and extinct inioids indicates that in South America, a patchwork pattern of freshwater invasion of river basins occurred, paralleling the recently reported occurrence of a platanistid from the Peruvian Amazon (Miocene; Bianucci et al., 2013) and the possible freshwater invasion of California's San Joaquin Valley by Parapontoporia (Boessenecker and Poust 2015, see above) - indicating significant adaptability among modern and extinct river dolphins and their marine relatives. 

The early evolution of baleen whales is now revealed by a series of nice transitional fossils including many well-preserved skulls showing a gradual evolution of baleen whale "features" from the ancient basilosaurid skeletal plan. However, most early fossils of odontocetes - toothed whales - are already highly derived, quite obviously being early dolphins with facial features consistent with echolocation and already well-telescoped skulls (telescoping is the posterior migration of rostral bones over the braincase, thought to "track" the posterior migration of the bony nares). Some putative latest Eocene odontocetes are reported from the Olympic Peninsula of Washington, but they are not yet really published (a short conference paper exists - not just an abstract - but there are no figures, nothing is named, and a longer paper is promised). A new fossil dolphin from the Ashley Formation here in Charleston, appropriately named in this paper as Ashleycetus planicapitis, is noted for its rather plesiomorphic (="primitive morphology") skull including limited telescoping, anteriorly positioned bony nares, and maxillae that only cover part of the frontal. The top of the skull is remarkably flat when viewed from the side, hence the species name. With the exception of the enigmatic dolphin Archaeodelphis (the locality of which is unknown, but thought by some to be from the Oligocene Ashley or Tiger Leap Formation of the Charleston area), Ashleycetus is the most plesiomorphic known odontocete - and given that much more derived dolphins are known from the Oligocene Ashley Formation, it's already a relict species. Given what we know of the Oligocene cetacean record, I would expect Ashleycetus-like odontocetes to show up in much earlier rocks, perhaps somewhat more archaic forms even in the late Eocene. This paper is absolutely overflowing with ideas and good observations, and given my new job working with early odontocetes here in SC - I've been finding myself re-reading parts of this paper over and over. This paper also provides some new photographs, observations, and interpretations of Xenorophus sloani (another Oligocene dolphin from the Ashley Formation of Charleston) and Mirocetus riabinini, a weird odontocete from the Oligocene of Azerbaijan originally mistaken for a toothed baleen whale (an assumption or cognitive bias which I would wager was originally based on its size). Students of early odontocete evolution will want to read this paper very carefully. Lastly, as an aside, an interesting talk at SVP by my colleague Jorge Velez-Juarbe (LACM) included a cladistic analysis that recovered Ashleycetus as a basal member of the Simocetidae - a group of Agorophius-like dolphins with downturned snouts.

This is another review article, so this will be brief. Most of this review is concerned with the ecology and evolution of the invertebrate fauna inhabiting modern whale falls - which, for the uninitiated, are whale carcasses that have sunk down to the deep sea floor and host a very distinctive fauna of marine invertebrates also seen at deep see vents and methane seeps. Occasionally fossil cetaceans have been recovered with trace fossils or associated/attached body fossils of these deep sea specialists, such as vesicomyid clams and the bone eating worm Osedax (which doesn't have a mineralized skeleton, but produces distinctive borings in whale bones which do preserve). This summary concludes that most elements of modern whale fall communities had their origins during the Oligocene, corresponding to the diversification of the Neoceti - one third of all extant genera of cold seep mollusks appear during the late Eocene and early Oligocene, tightly corresponding to the diversification and worldwide dispersal of Pelagiceti (Neoceti + Basilosauridae). These authors suggest that more research into the evolution of bone lipids (the principal source of nutrients for many whale fall specialist invertebrates) within extinct cetaceans should be conducted - which may perhaps be inferred from postcranial bone histology. Lastly, modern evidence of the "reef stage" - the fourth stage in the evolution of a single whale fall (after the mobile scavenger, enrichment-opportunist, and sulfophilic stages) has been criticized by other whale fall biologists, yet has support from fossils. This stage was hypothesized to exist as a period after which the nutrients have been completely removed from the bone, but because the bones still physically extend above the seafloor sessile filter feeding invertebrates will colonize the bone to take advantage of a higher current. Many examples of barnacles, serpulid worms, bryozoans, and other sessile filter feeders are known from marine vertebrate skeletons preserved in deep marine settings.

Basilosaurus is known from two species from eastern North America (B. cetoides) and northern Africa (B. isis) and represents the largest basilosaurid archaeocetes known - giant serpentine whales with quasi-vestigial hindlimbs that lived during the late Eocene (a third possible species is debated but has been reported from Pakistan - B. drazindai). All basilosaurids are characterized by fearsome dentitions with caniniform anterior teeth and large, triangular, cuspate shearing cheek teeth - and like most archaeocetes, have tiny braincases with enormous jaw muscle attachments. Smaller archaeocetes from the same deposits as B. isis, including several juvenile skulls of the small basilosaurid Dorudon atrox, have been found with large tooth punctures, reasonably hypothesized by Julia Fahlke to be tooth punctures from B. isis. Does Basilosaurus have sufficient bite force to cause bite marks like this? What is the bite force of an archaeocete? We can't go out and put a force gauge into the mouth of an extinct organism - so computer modeling, specifically finite element modeling -  provides a means by which to estimate bite force. I won't go into how FEM modeling works, principally because I am not mentally equipped to do so - but it can be done using CT data or, as in this case, a 3D surface scan of a 3D object. Using high resolution CT data permits density/strength values to be placed onto tiny 3D 'cells' (voxels: aka 3D pixels) with the density derived directly from the CT scan. Another method is to use a surface scan and arbitrarily assign bone wall thickness (or treat it as a solid) and uniform bone density/strength within. When scaled to the same size, FEM indicates that Basilosaurus had comparable bite forces with giant predatory pliosaurs (e.g. Pliosaurus kevani). Although lesser in magnitude than the highest bite forces measured and estimated for large crocodylians and dinosaurs, bite forces predicted (16,400 newtons) at the upper third premolar (P3) of Basilosaurus exceed those of any other mammal and additionally exceed predictions of force as expected from its relatively narrow skull. Notably, Basilosaurus was capable of comparably higher bite force at the tip of its snout than crocodylians. Bite force is indeed consistent with indenting and breaking bones, and feeding behavior likely consisted of catching prey with the anterior teeth and mastication (or should we say in this case, "chopping") of prey items with the cheekteeth. 

This paper utilizes newly recovered molecular data from the extinct Steller's sea cow (see Estes et al., above) to run a comprehensive phylogenetic analysis - combined with morphological data - of modern and extinct sirenians. Much of this paper is concerned with new molecular results - which are interesting, but this post is focusing on paleontological advances so I'll focus on those. The second sentence of the abstract goes like this: "The phylogenetic affinities of [Hydrodamalis gigas] to other members of this clade, living and extinct, are uncertain based on previous morphological and molecular studies." This raised huge red flags for me because from everything I had read by two of the foremost experts in the world on sirenian evolution and anatomy, Daryl Domning and Jorge Velez-Juarbe - had indicated that Hydrodamalis is closely related to the southwestern Pacific Dugong dugon, and that giant hydrodamaline sea cows evolved in situ within the north Pacific during the late Neogene, giving rise to Hydrodamalis by the late Miocene/early Pliocene - all of this appeared non-controversial. The paper of course reports similar results, and both Daryl and Jorge are coauthors - upon further reading, other researchers have produced some rather odd results with the west Indian manatee coming out as more closely related to Hydrodamalis - which doesn't make sense for a number of reasons, such as the shared presence of tail flukes. The article also provides a brief but handy summary of macroevolutionary trends in sirenian evolution.


The incompletely preserved dolphin Prosqualodon marplesi was named in the 1960's from the upper Oligocene-lower Miocene Otekaike Limestone of New Zealand, originally placed in the squalodontid genus Prosqualodon. R.E. Fordyce recognized how dissimilar it was to Prosqualodon, and placed it in the squalodelphinid genus Notocetus instead when he described the other NZ dolphin Waipatia maerewhenua. Last year, my labmate Yoshi Tanaka published a reevaluation of "P." marplesi and assigned it to the new genus Otekaikea - and surprisingly recovered this specimen in a cladistic analysis as a sister taxon to Waipatia, and reclassified Otekaikea marplesi as a waipatiid. In this new paper, Tanaka and Fordyce describe a second species, Otekaikea huata, based on a much more complete specimen. Otekaikea huata has a similar braincase and earbones, differing in only a few subtle ways obvious only to students of whale anatomy too nuanced to repeat here. However, the relatively complete holotype specimen of O. huata exhibits many notable features that are interesting from a functional perspective. The rostrum is very elongate, and the teeth are nearly homodont posteriorly, with simple crowns and single roots - and transition anteriorly into tusklike apical teeth. The anteriormost tooth is huge, about 4-5 inches long, and straight - the tusk would have been procumbent, and probably not functioning for biting prey. The facial region is strongly dish-shaped, indicating the presence of a melon and associated facial muscles involved in sound production - clearly indicating that Otekaikea huata used echolocation. Hearing was specialized like many modern odontocetes, with large sinuses in place around the earbones, either for soft tissues or pneumatic sinuses - acoustically isolating the inner ear from bone-conducted sounds in the skull. Interestingly, most Oligocene odontocetes to date are known for their comparably modest rostral proportions - whereas nearly all non-squalodontid odontocetes from the early Miocene have embarrassingly elongate snouts, like Otekaikea huata (and nobody has really offered a good solution as to why). Otekaikea thus may represent the first known member of this functional group of longirostrine dolphins, giving a preview of future affairs.


During the late 19th and early 20th centuries a number of fragmentary but anatomically curious fossil cetaceans and penguins were discovered and named from various Oligocene marine rocks in the Waitaki Valley region of the South Island of New Zealand. Several other papers summarized above have also dealt with some of these historical specimens (Boessenecker and Fordyce 2015, Tokarahia), as well as more recently collected fossils (Boessenecker and Fordyce, 2015 – Waharoa; Tanaka and Fordyce, 2015, Otekaikea huata; Tsai and Fordyce, 2015, Horopeta). Early identifications and efforts to properly interpret these fossils were hampered by their incompleteness and lack of comparable material. The species Microcetus hectori is one of these, based on a fragmentary mandible and some isolated teeth collected from the upper Oligocene Otekaike Limestone in 1881 by notable geologist Alexander McKay. The teeth are tiny with high crowns, accessory cusps on the posterior side, and labial and lingual cingulum – in person, I call them “cute” (Yoshi Tanaka had these on the desk next to me in my office for several months while working on this chapter of his thesis). This fragmentary specimen was originally placed in the genus Microcetus based on its inferred dental similarity with Microcetus ambiguus; however, detailed observations show that the teeth differ in many regards, and at a gross level are more similar with other NZ dolphins like Waipatia. A skull in a block of sediment was also collected and discovered over 100 years later by R.E. Fordyce. Given that numerous teeth, a partial mandible, and most of a braincase were now known, the authors included it within a cladistics analysis – wherein it was allied with Waipatia maerewhenua. The authors recombined it as Waipatia hectori. This, with the reinterpretation of “Prosqualodonmarplesi as the new genus of waipatiid Otekaikea and the naming of a second species, Otekaikea huata (see above), really shakes up what was formerly thought of in terms of odontocete diversity in New Zealand as many of these seemingly different odontocetes were formerly thought to represent other odontocete families. And, there are more waipatiids to come!


Readers of this blog already know I have a healthy obsession with fossil walruses. Avid readers of this blog remember my series of posts on walrus evolution and already know that most extinct walruses did not have tusks, and that tusks only really characterize a few highly derived walruses from the Pliocene and younger. Most fossil walruses had skulls (and likely outward appearances in life) broadly similar to sea lions. After all, walruses are pinnipeds, and no other modern pinnipeds have tusks or such highly specialized diets - so we should expect the modern walrus to have evolved from a more generalized, fish eating ancestor. Fossils from the North Pacific - chiefly California and Japan - now illuminate the early history of walrus evolution stretching back to the early Miocene and include many early diverging, sea lion-like "imagotariine" walruses like Neotherium, Imagotaria, Prototaria, and Proneotherium. In California we have the small-bodied Neotherium in the mid Miocene (~15 Ma) and by the late Miocene (~9-10 Ma) we have the much larger Imagotaria downsi, with simpler teeth (less cusps, single rooted teeth). In 2006, Naoki Kohno reported Pseudotaria muramotoi from the early late Miocene of Japan (9.5-10 Ma), which is intermediate in terms of morphology and size between Neotherium and Imagotaria. This new paper stems from Yoshi Tanaka's master's thesis research in Japan and names a new imagotariine from the same deposits (Ichibangawa Formation, Hokkaido) that is quite a bit more complete than Pseudotaria. The type specimen of the new species Archaeodobenus akamatsui is broadly similar to Pseudotaria but differs in many features of the basicranium as well as the morphology of the cervical vertebrae. Only the left side of the skull is preserved, but hyoid apparatus and mandible as well as many teeth were recovered; additionally most of the cervical and thoracic vertebrae were recovered as well as ribs, a sternebra, scapula, and humerus. This specimen demonstrates that two similar walruses coexisted in the late Miocene of Japan, which is interesting - but multispecies walrus assemblages are already known from the Pliocene San Diego Formation (Valenictus, Dusignathus, Odobenini indet.) and mid Miocene Sharktooth Hill Bonebed (Neotherium, Pelagiarctos) of California, so this is perhaps not very surprising.

This study marks another contribution by retired fossil preparator Howell Thomas into the field of paleopathology - the study of disease in the fossil record. This study surveys osteochondrosis in modern and fossil marine mammals. Osteochondrosis has an idiopathic origin - idiopathic roughly translates to "we have no idea what exactly causes it." Osteochondrosis usually comprises damage to the articular surface of a long bone, and is thought to be caused by trauma to the joint, like extreme vertical loading of a human knee; shear loading, avulsions, and continued low-grade traumas to the same location along with some other problems can cause osteochondrosis. Trauma upsets normal cartilage growth at the joint and bone death occurs below the cartilage which manifests as a deep pit on the articular end of the bone. The authors figure and describe pits in humeri, ulnae, scapulae, and skulls of extant marine mammals including walrus, monk seals, and narwhals. Osteochondrosis is present in postcranial bones of the desmatophocid pinnipeds Allodesmus "kelloggi", Allodesmus kernensis, the "cetothere" Tiphyocetus temblorensis, a skull of the sperm whale Aulophyseter morricei, the atlas vertebra of the dolphin Zarhinocetus errabundus, and postcrania of isolated odontocetes and the desmostylian Neoparadoxia cecilialina from the Monterey Formation. This study reports the first occurrences of osteochondrosis both within modern and fossil marine mammals. It was not found in any sirenians, but instead was found only within amphibious pinnipeds and desmostylians (which could become injured when exiting/entering the water) and cetaceans (which could become injured when breaching or similar behavior).

Modern baleen whales are readily identifiable based upon their baleen as well as their enormous body size; indeed, their great mass is perhaps what best captures the imagination of the public. As alluded to above, baleen whales had rather humble beginnings – the “chonecetine” aetiocetids (e.g. Chonecetus, Fucaia) were scarcely larger than a harbor porpoise (~2 meters long). Other aetiocetids, like Aetiocetus and Morawanacetus, reported from Japan and the Pacific Northwest – are only slightly larger, perhaps approaching the size of a large bottlenose dolphin (~3-4 meters). This rather small range of body sizes contrasts with the somewhat larger (and contemporaneous) early baleen-bearing eomysticetids, which were about the size of minke whales (5-8 meters). A new aetiocetid fossil from the upper Oligocene of Hokkaido (northernmost major island of Japan) reported by Tsai and Ando consists of a squamosal and a periotic similar in morphology to Morawanacetus yabukii – but is approximately twice as large, with a body length estimate of 8 meters. This body length is in the size range of eomysticetids, and expands the range of size disparity amongst toothed mysticetes. Furthermore, because this large morawanacetine is found in the same deposits as smaller Morawanacetus yabukii, some degree of niche partitioning must have been present. Future finds preserving the feeding apparatus of the large, unnamed morawanacetine may reveal how niche partitioning occurred.

 One of the earliest fossil baleen whale earbones I ever saw photos of was a specimen collected by Ron Bushell, formerly of Eureka in northern California, who had collected it from the Plio-Pleistocene Rio Dell Formation nearby in Humboldt County. As a high school student interested in local paleontology, it boggled my mind that nobody could identify it. Years later I found out that it, and other neat specimens collected by Ron, had been kindly donated to Sierra College in Rocklin, CA. After spending a couple years during my Ph.D. staring at mysticete earbones until my eyes felt like they were going to bleed, I realized it was probably an early record of a gray whale – so I invited my labmate Cheng-Hsiu Tsai to describe it. Turns out it’s nearly identical to modern Eschrichtius robustus, so we identified it as Eschrichtius sp., cf. E. robustus. This specimen, consisting of a tympanic bulla and a compound posterior process, is more similar to modern E. robustus than a Pliocene specimen published in 2006 from Japan identified as Eschrichtius sp. As it happens, Bushell’s specimen is from the uppermost Rio Dell Formation, making it early Pleistocene (~1-2 Ma) in age – a time period nearly completely unrepresented for marine mammal fossils in the east Pacific. There is a more “primitive” unnamed genus of gray whale (Eschrichtiidae) from older Pliocene rocks in California, but no bona fide records of Eschrichtius; the Pliocene of California is probably well sampled enough to declare that Eschrichtius was not present (but at least a half dozen other mysticetes were present instead). Given the delayed occurrence of Eschrichtius in California relative to Japan, we hypothesized that the modern gray whale evolved in the western North Pacific during the Pliocene, and dispersed to the eastern North Pacific during the early Pleistocene – sometime after the Plio-Pleistocene marine mammal extinction which led to the demise of eastern Pacific walruses (Dusignathus, Valenictus), the bizarre porpoise Semirostrum, and other cetaceans. 

 The idea of ancestor-descendant relationships has pervaded paleontology since the 19th century, but with the advent of cladistics and the emphasis on phylogenetic relationships a bizarre misconception that we cannot identify ancestors and descendants in the rock record has arisen. Certainly this is an artifact caused by the fact that cladistics - the dominant method for inferring phylogenetic relationships amongst modern and extinct organisms - can only infer "relatedness" but not time. Thus, inability to interpret ancestors versus descendants is based on a limitation of our current methodological paradigm - a limitation that this new study seeks to circumvent. This study investigates the highly problematic and controversial relationships of the pygmy right whale, Caperea marginata. A very Caperea-like fossil, Miocaperea pulchra, is known from the late Miocene of Peru (and these authors suggest that it could even be recombined as Caperea pulchra, given the similarity). A well-known but underappreciated aspect of anatomy is that growth of vertebrates roughly parallels the evolutionary history - in an imperfect sense, not quite as predicted by Ernst Haeckel (ontogeny recapitulates ontogeny). Generally speaking, in many vertebrate groups, juveniles will look like ancestors - to the point where juvenile hadrosaur dinosaurs have been misinterpreted as small adults of late surviving archaic hadrosauroids. An earlier study by some Canadians and my dear friend Liz Freedman-Fowler (Hi Liz!) found that when juveniles of known hadrosaur species within different families were coded into an existing cladistic dataset, the juveniles all plotted together on the paraphyletic "stem" of the group. This concept applies to cetaceans as well. In this study, Miocaperea, adult Caperea, and juvenile Caperea were coded as different OTUs into two existing cladistic matrices. In both cases, Miocaperea was phylogenetically bracketed between juvenile and adult Caperea. Given this, and relatively slow change in the neobalaenid lineage and neoteny within the ontogeny of modern Caperea, Miocaperea and Caperea could be end-members of a late Miocene-Holocene anagenetic lineage undergoing evolutionary stasis. Ultimately, this does raise additional red flags for interpreting the relationships of cetaceans based on juvenile specimens (e.g. Nannocetus eremus, Parietobalaena palmeri).

Lunge feeding - otherwise known as gulp feeding - is one of the more derived means by which baleen whales filter feed for prey. As discussed above (see Boessenecker and Fordyce 2015: Waharoa) skim feeding consists of swimming slowly through the water column and continuously filtering out planktonic prey - this is utilized by modern right whales, probably Caperea, and inferred in eomysticetids. Gray whales feed by ingesting large volumes of sediment and filtering out small benthic crustaceans. Rorquals (humpbacks, blue, fin, minke whales) lunge feed - they swim fast towards prey and rapidly open the mouth and close it; water is expelled by the slowly contracting throat pouch. Many fossil baleen whales from the Oligocene of New Zealand - particularly the Duntroonian stage (27-25 Ma) - are eomysticetids, but by the Waitakian (25-23 Ma) are much more rare, and early "Kelloggithere" like mysticetes are present - these are poorly known, poorly understood whales like Parietobalaena; Mauicetus parki from the Waitakian (Otekaike Limestone, Milburn Limestone) of NZ is a prime example. Their relationships are unclear, and do not belong within the true Cetotheriidae, and a number of other families have been proposed. This new study reports perhaps the oldest member of this grade, Horopeta umarere, from the transition between the Kokoamu Greensand and the Otekaike Limestone in south Canterbury, New Zealand (same locality as one of the juvenile specimens of Waharoa ruwhenua). This whale has a partial skull that was disarticulated and bioeroded and thus many of the bones do not articulate, but the braincase is otherwise well preserved and includes immaculately preserved earbones - which are weird looking. They resemble the younger Mauicetus parki, but differ from pretty much all other Chaeomysticeti (except the more archaic eomysticetids) in lacking fusion of the posterior process of the earbones, indicating rather archaic status amongst the mysticetes. The mandibles are huge with a wide cross section and - most importantly - are laterally bowed like a humpback whale. These mandibular features are consistent with lunge/gulp feeding, and represent the geochronologically earliest occurrence of such adaptations. Horopeta also has a rather large, robust sternum with attachment points for multiple ribs - and is thus more primitive than the delicate platelike sternum of eomysticetids like Waharoa and Tokarahia. More strange mysticetes have yet to be described from the Oligocene of New Zealand (and Washington, U.S.A.) and will certainly complicate the emerging picture of early mysticete evolution.

            Elephant seals (Mirounga spp.) are the largest members of Carnivora and the most sexually dimorphic of all mammals, with males weighing and measuring many times larger/longer than females and having extreme ritualized behavior and bizarre probosces for display. Elephant seals live in the Antarctic and Southern Ocean as well as the eastern North Pacific. Despite having well-studied ecology and behavior, virtually nothing is known of their evolution. Bits and pieces have been mentioned, but have never been described until this new paper by Ana Valenzuela and others on middle-late Pleistocene records of Mirounga. These fossils include skull fragments and a partial mandible and some other fragments from Mejillones in Chile, and represents the first described fossil record of elephant seals. The fossils are not very old – and most other undescribed records of Mirounga are also Pleistocene, suggestive of a geochronologically shallow history of elephant seals.

Fossil pinnipeds are widely reported from the Miocene and Pliocene of South America, but most of the fossils - unlike today - are of phocid seals (aka earless or true seals) - rather than the otariid fur seals and sea lions that currently inhabit these coasts. Included within the well-sampled Pisco Formation of Peru are the well-known seals Acrophoca and Piscophoca - Piscophoca is a generalized and monk seal-like, whereas Acrophoca has an unusually elongate, narrow skull. Both of these have also been reported from the Bahia Inglesa locality in Chile. In 2012, an additional phocid seal similar to Piscophoca but with enlarged cheek teeth (and therefore with a possible durophagous diet) was named Hadrokirus. This new paper by Ana Valenzuela-Toro and colleagues reports yet another phocid seal - Australophoca changorum - from the Pisco Formation, but this seal is tiny and approximately the size of extant ringed and Baikal seals (Pusa). Australophoca is known mostly by postcrania (humerus, radius, innominate, femur, calcaneum, astragalus) and is too incomplete to be coded into a phylogenetic analysis, but it is probably a monachine seal (southern seal) rather than a phocine (northern seal) based on an elongate deltopectoral crest of the humerus, and lacking an entepicondylar foramen in the humerus. Also, nearly all southern hemisphere fossil phocids are monachines (one notable possible exception is Kawas benegesorum from Argentina) so this is hardly surprising. The epiphyses of all specimens of Australophoca are completely fused, indicating adult status, and additional specimens are recorded from Bahia Inglesa in Chile. The tiny adult is surprising, as most southern hemisphere pinnipeds are quite large, with the exception of the Juan Fernandez and Galapagos fur seals (Arctocephalus phillippii and Arctocephalus galapagoensis, respectively) which curiously inhabit the west coast of South America today. These authors point out that the range in body size observed amongst Mio-Pliocene phocid assemblages from Peru and Chile is comparable to that observed today in Alaskan waters. Sometime during the late Pliocene or Pleistocene this diverse phocid assemblage went extinct and was replaced by a roughly modern assemblage by at least the middle or late Pleistocene judging from fragmentary fossils of younger age - perhaps owing to changes in upwelling, coastal uplift, and changes in coastal geomorphology.

Velez-Juarbe, J., and D.P. Domning. 2015. Fossil Sirenia of the West Atlantic and Caribbean region XI. Callistosiren boriquensis, gen. et sp. nov. Journal of Vertebrate Paleontology 35:1:e885034.

This study is the eleventh (!!!) installment of the series of papers dedicated to fossil sirenians from the west Atlantic and Caribbean, started by reknowned sireniologist Darly Domning - the last few contributions (9-11) have been coauthored by Jorge Velez-Juarbe, and include some spectacular finds from South Carolina, Florida, and Jorge's home territory - Puerto Rico. The holotype of Callistosiren is an impressive skull collected by Jorge back in 2005 - and it has made some appearances on his blog and on SVP posters. It's a large medium sized dugongid from the Oligocene Lares Limestone of Puerto Rico, characterized by mild rostral deflection (nowhere near as vertical as extant Dugong, but not quite as horizontal as the giant hydrodamalines I'm used to in the north Pacific) and large tusks with enamel present only on the medial surface of the tusk. Notably the ribs and vertebrae show substantially less dense bone than other sirenians. This new discovery highlights how diverse the dugongid lineage was in the Oligo-Miocene of the Atlantic and Caribbean basins; in the North Pacific, there tends to be only one or two sirenians present (possibly owing to competition with desmostylians?) whereas earlier work by Jorge has already demonstrated that many Atlantic, Caribbean, and Indian ocean sirenian faunas are characterized by multispecies assemblages with evident niche partitioning. Callistosiren is the first sirenian recorded from the late Oligocene (Eocene and early Oligocene examples are already known - e.g., Pezosiren, Priscosiren), and an undescribed halitheriine dugongid is also known from the coeval Mucabarones Sand in Puerto Rico. Rostral deflection and tusk size indicate that Callistosiren likely fed on rhizomes of relatively large seagrasses, and the authors further hypothesize that the low density of postcrania - virtually unknown in all other post-Eocene sirenians - may be an adaptation for deeper diving and foraging at greater depths.

Sperm whales are some of the largest animals to have ever lived, and the largest non-filter feeding predators in the world. The public, however, is generally only familiar with the giant sperm whale Physeter - yet two other tiny sperm whales, the rare pygmy and dwarf sperm whales (Kogia spp.) are quite fascinating in their own right. Kogia, placed in its own family Kogiidae by most contemporary cetologists, shares several features with the much larger Physeter such as a convex toothless palate, a supracranial basin, extreme cranial asymmetry, lower teeth lacking enamel, and an elongate mandibular symphysis. Kogiids and physeterids share a common origin someplace around the middle or early Miocene (judging from the age of the oldest known crown physeteroid, Aulophyseter morricei from the Sharktooth Hill Bonebed of California). This study reports a new kogiid, Nanokogia isthmia, from the upper Miocene Chagres Formation of Panama. The first specimen discovered, a referred braincase, was collected by the lead author Jorge Velez-Juarbe while a postdoc student in the Panama Canal Project. Other fossil kogiids are somewhat larger than extant Kogia, such as Scaphokogia (late Miocene, Peru) and Aprixokogia (early Pliocene, North Carolina). The skull of Nanokogia is somewhat smaller and similar in size to extant Kogia, but differs in having a narrower braincase and somewhat more elongate rostrum. Unlike Aprixokogia and many other fossil physeteroids, Nanokogia shares with extant Kogia and Scaphokogia a lack of upper teeth. Nanokogia is most similar to Praekogia cedrosensis (late Miocene, Baja California) and extant Kogia, and plots out as closely related to each in the cladistic analysis. Modern Kogia has a much smaller spermaceti organ than Physeter (the spermaceti organ dorsally overlies the melon - these two organs make up the classic soft tissue "forehead" of Physeter). An enlarged premaxillary sac fossa in Nanokogia, like Praekogia and Scaphokogia, seems to indicate these extinct dwarf sperm whales retained a well-developed spermaceti organ - but lost or reduced in Kogia and the early-mid Miocene kogiid Thalassocetus. Lastly, the authors point out that during the late Miocene, a much higher level of cranial disparity involving the feeding apparatus is apparent amongst physeteroids (paralleling other groups of late Miocene-Pliocene cetaceans as well).

This is one of the strangest papers on fossil marine mammals this year, and I do not mean that in a bad way - it's really a good example of thinking outside the box when it comes to applying fossil vertebrates towards answering questions outside the realm of vertebrate paleontology. It all begins in 1964, when a young undergraduate student from Yale interested in paleoanthropology was on a field expedition with Bryan Patterson in the Turkana Basin in Kenya and found what everyone assumed to be a weird turtle shell. Later on, it was prepared and discovered to be a cetacean - and not only that, but a rare beaked whale (Ziphiidae). Mead's discovery "derailed" his future in paleoanthropology and drove him towards studying cetaceans - he produced a spectacular dissertation on dissecting out the facial region of modern odontocetes in order to investigate the source of echolocation-related sound production, and quickly became an expert on the anatomy and biology of beaked whales. One of his first papers (Mead, 1975) focused on the Turkana ziphiid. Then, some years later, the specimen went missing, and wasn't rediscovered until somebody cleaned out Stephen J. Gould's old office at Harvard in 2011, which was temporarily being used for storage. The importance of this specimen actually lies in its geologic context - the Turkana Basin is entirely terrestrial, and Mead speculated that it was an individual that swam up a river and became stranded. Wichura et al. use this in conjunction with data on how far modern oceanic cetaceans have swam up rivers to put a maximum elevation of about 30 meters above sea level on this fossil at the time, indicating it must have swam 600-700 km from the hypothesized shoreline at the time. During the early middle Miocene, at the time of the stranding, the coastal plain in this area consisted entirely of tropical rainforest with significant rainfall. Sometime after, the entire region began to uplift and it became very arid - leading to the first savannas in east Africa, an event thought to have driven the earliest human ancestors (e.g. Ardipithecus, australopithecine hominins) from the safety of the forest and onto the plains. Timing of this uplift has been poorly constrained, and the occurrence of this ziphiid so far inland now indicates that uplift must have taken place sometime after 17 Ma.