Wednesday, December 24, 2014

2014 in review: Advances in marine mammal paleontology

Merry Christmas/Happy Hannukah/Chrimbus/Festivus! Welcome to the third annual review of advances in marine mammal paleontology! There were over 50 new publications in marine mammal paleontology this year, and this took quite a while to work up; about 1/3 of the papers I didn’t even really get a chance to read until this winter (er, summer). As usual, this list is supposed to be comprehensive, so if I have missed something, please let me know – hopefully I’ll have time to update it. I may not, because ultimately my thesis is more important. As in the past two years, I include papers that have been published since January 2014 – however, for papers that came out online in 2013 and were only included in published journal issues in 2014 – these will remain in last year’s post (and the same goes for some of the following studies, some of which will almost certainly have permanent 2015 citation dates despite online publication in late 2014). There are three or four I still have yet to add, so stay posted and check again in a few days.

It does bug me when I see new papers that have left out citations to important new research; I try to devote some time to reading new publications as they come out, and hopefully with these posts statements like “well I didn’t really know about such and such article, I don’t know how to use google” or “I’ve literally been living under a rock” will no longer be decent excuses. Anyway, I hope this huge review will be of service for other marine paleomammalogists!

This paper formed one of the core chapters of former labmate Gabriel Aguirre-Fernandez’ doctoral thesis, and reports a well-preserved and beautifully prepared odontocete skull from the lower Miocene Kaipuke Siltstone in northwest Nelson, South Island, NZ. This specimen was for years thought to be the earliest record of kentriodontids (but keep reading). The skull is rather small, perhaps similar in size to a Hector’s dolphin or harbor porpoise. It appears to have had a polydont single rooted dentition and a subtriangular, acutely pointed rostrum and convex palate – overall not terribly different from a dwarfed Waipatia aside from the tooth rooting. Cladistic analysis indicated rather strongly that this is no kentriodontid, and is in fact a stem odontocete of uncertain position; it’s more derived than xenorophids (see Geisler et al., below, for more on xenorophids), but more primitive than beaked whales and sperm whales. They named the new dolphin Papahu taitapu, meaning dolphin from Te Tai Tapu (the area of Nelson where it was collected from) in Māori.

One of the most profound and strangest discoveries in marine mammal paleontology is the discovery of aquatic adapted sloths from the Miocene and Pliocene of Peru (and now Chile as well). These were first named Thalassocnus in 1995, and in the following decade a number of additional species and a ton of skeletons were reported and described. Common questions I get about these from colleagues in non-mammalogical fields of paleontology are “How do we know they’re actually aquatic and not just terrestrial sloths that floated out to sea?”. It’s an excellent question that can be addressed taphonomically and adaptationally. Taphonomically speaking, these skeletons have only been found in marine sediments and none yet from terrestrial deposits – which have other types of sloths. Further, apparently much of the coastline of Peru during the late Neogene was inhospitable desert. These sloths also have modified limb proportions and large muscle attachment areas on caudal vertebrae for undulation of the tail (if I recall correctly; it’s been almost ten years since I last read the original studies). Another approach – taken by the new article by Eli Amson - is paleohistology. Most tetrapods that return to the water go through a period where postcranial bones become denser, as a buoyancy adaptation thought to offset positively buoyant air-filled lungs during shallow diving. This pattern of increased bone density (pachyostosis – increased cortical bone thickness; osteosclerosis – decreased medullary cavity volume; and pachyosteosclerosis – combination of both) characterizes the early evolution of most marine tetrapod groups, including (but not limited to) ichthyosaurs, mosasaurs, pachypleurosaurs, placodonts, most flightless diving birds (penguins, mancalline auks), pinnipeds, sirenians, cetaceans, desmostylians, and sea otters. These patterns of bone thickness are to a degree obvious externally when comparing with terrestrial ancestors, particularly with modern specimens where differences in bone mass and density are easily felt during handling. In fact, using Archimedes principle is an easy way to determine bulk bone density for modern specimens – but for fossils, there is always an unknown mass of sediment and cement inside, and the vagaries of diagenesis otherwise make this approach impossible. Instead, we can either chop up bones and look at them under a microscope – or use CT scanning to get a cross section. In either case a software program called Bone Profiler can be used to quantitatively compare cortical bone thickness and a dimensionless compactness metric across taxa. Amson et al. did just this with ribs, femora, and tibiae of three species of Thalassocnus (T. antiquus, T. natans, T. littoralis, and T. carolomartini) and three outgroups (an anteater, an extant sloth, and the extinct ground sloth Hapalops) and found that bone density does in fact increase from the basal end of the tree to the end. The three species of Thalassocnus are stratigraphically separated and perhaps thought to represent cross-sections of a lineage through time, and interestingly (and predictably) bone thickness increases in younger species. As an aside, some groups that exhale and let their lungs collapse – an adaptation for deep diving – have actually gone the other way around, and evolved light porous bone (osteoporotic – yes, the very same condition osteoporosis that disproportionably affects women) to offset the loss of buoyant air-filled lungs, and includes derived ichthyosaurs, most modern cetaceans (particularly odontocetes), elephant seals, and a single desmostylian (Desmostylus). I’ve taken a bunch of thin sections of eomysticetid ribs, and have tinkered around with converting images of my thin sections for analysis in Bone Profiler (needs more tinkering still).

Egypt and Pakistan are perhaps the two countries most famous for producing spectacular fossils of archaeocete whales that have provided us with the skeletal evidence demonstrating the famous land to sea transition ("feet to flippers") that is now a poster child for vertebrate macroevolution. Although often outshined by Pakistan, India has produced a number of important fossils as well, including pakicetids, remingtonocetids, and protocetids; unsurprisingly, most of these fossils come from Kutch, the westernmost part of India that is fairly close to Karachi, the capital of Pakistan. This new study by Bajpai and Thewissen reports several new protocetid specimens from India. Some of these include a new partial mandible and a maxilla fragment with teeth, the first for the earlier described protocetid Indocetus rahmani. They also describe two new genera of protocetids; the first, Kharodacetus sahnii, is represented by a well-preserved anterior skull and partial mandibles with well-preserved teeth, and a few referred cranial specimens. Kharodacetus is one of the largest known protocetids worldwide, and is similar in size to the basilosaurid Zygorhiza - heralding the transition from medium-sized protocetids to the larger body sizes attained by later basilosaurids. The second new genus is based upon a nearly complete but poorly preserved skull, and is named Dhedacetus hyaeni (after the hyenas that live at the type locality!). Dhedacetus also has a referred vertebral column, and this column indicates it had a tail with robust musculature, indicating tail-based propulsion (rather than hindlimb propulsion as proposed for Georgiacetus by Mark Uhen a few years ago). Bajpai and Thewissen further point to Maiacetus (one of the only completely known protocetids) and suggest it may have actually had a caudal peduncle and fluke (although see Uhen 2014 on Natchitochia, below).

The early Miocene squalodelphinid dolphin Notocetus vanbenedeni was originally reported from Argentina. Notocetus vanbenedeni is a long-snouted archaic dolphin, and the squalodelphinids are a group of platanistoids – a formerly diverse group of odontocetes more closely related to the modern ganges river dolphin, Platanista. Another platanistoid – Otekaikea marplesi – was formerly placed in Notocetus by my adviser Ewan Fordyce in the 1990’s, but a redescription of that odontocete indicates it’s more closely related to Waipatia (see below: Tanaka and Fordyce). Other squalodelphinids include Squalodelphis (Italy), and most importantly, Huaridelphis (see Lambert et al., below). Huaridelphis was also collected from the Chilcatay Formation, indicating that two squalodelphinids were likely sympatric. Notocetus vanbenedeni is now known from both the west and east coast of South America during the early Miocene.

As discussed elsewhere below, there has been an explosion of research on the morphology, phylogeny, and feeding ecology of true cetotheres, the Cetotheriidae sensu stricto. One such new cetothere, one that is obviously closely related to my favorite cetacean, the dwarf baleen whale Herpetocetus, is reported in this new paper by Michelangelo Bisconti and is named Herentalia nigra. It’s based on Herpetocetus-like braincase from the upper Miocene of Belgium, and differs principally from Herpetocetus in only a few features – namely shorter ascending maxillae, longer exposure of the parietal at the “vertex”, a weird periotic, a squamosal cleft, and is much larger in absolute size. It is nearly identical to Piscobalaena nana from coeval strata in Peru, differing only in its larger size and presence of a squamosal cleft – which makes me wonder whether it might be a species of Piscobalaena. This study also criticizes the recent hypothesis that the pygmy right whale is an extant cetothere, even with a section of the discussion titled “Is Caperea marginata a cetotheriid sensu stricto?”. Bisconti brings up many points of disagreement on coding of various characters in the cladistic matrix, focusing on characters and character state definitions identified as supporting the relationship. Bisconti’s cladistic study instead supports the traditional view of balaenoid monophyly – supported by virtually all analyses aside from those by former labmate Felix Marx. I have my own reservations, but those will have to wait until my own enormous mysticete matrix from my doctoral thesis gets published (currently in review).

This paper was the first from my doctoral thesis to be published, and describes new (and old) eomysticetid material from the south island of NZ. The fragmentary skull and earbones of “Mauicetuswaitakiensis were never really properly described or figured, and more recent discoveries of the type species of Mauicetus (Mauicetus parki) reveal that “Mauicetuswaitakiensis belongs to a more primitive group of mysticetes. Its incompleteness, however, hampered any attempts to interpret it. A new skull with other associated bits (tympanoperiotic, mandible, vertebra, ribs) shared unique features of the posterior braincase and tympanic bulla of “Mauicetuswaitakiensis, convincingly linking the two together. So, we named the new genus Tohoraata raekohao for the more recently collected skull, and recombined the older species as Tohoraata waitakiensis. Tohoraata is Māori for “Dawn whale”, in parallel with the closely related “Eomysticetus”, which means the same thing. Cladistic analyses including Tohoraata will be published in a subsequent paper derived from my thesis research.

Bone-eating worms – Osedax – infest skeletons at modern whale whale falls on the deep sea floor, boring little holes into bone and digesting/dissolving bony tissues with the help of symbiotic bacteria. This activity leaves known traces in modern whale bones, and fossil whale bones from the Oligocene of Washington (USA) and Pliocene of Italy have already been found with Osedax traces. This study presents the first record of bone eating worm fossil traces from the Southern Hemisphere – which in and of itself is noteworthy, but not exactly groundbreaking. Rather, the traces we found on an Oligocene eomysticetid skull and mandibles are cross-cut with tooth marks from a shark or bony fish. Because Osedax generally colonize skeletonized bones (e.g. defleshed), these tooth marks are not attributable to scavenging of remnant soft tissues but instead reflect feeding upon the soft tissues of the worms themselves, which do leave a thin protective barrier of outermost bone – but generally too thin to dissuade vertebrate predators. Interestingly, modern ratfish and crabs have been observed ripping Osedax out of bones at whale falls.

Taphonomy is the study of fossil preservation, essentially including all processes that affect a dead organism from death until burial. It’s easy to study the actualistic taphonomy of terrestrial organisms, since those are environments accessible to us: we can watch a dead animal decompose our backyard, or chuck bones into a stream (or flume) and watch them get sorted. We’re less lucky with marine vertebrates, however; actualistic studies effectively include studies of carcasses and bones along shorelines, or studies of deep sea whale falls. These mark important contributions, but tell us virtually nothing about what happens in between on the continental shelf – a setting reflecting virtually all Cenozoic marine vertebrate assemblages. Instead, we can turn to the rock record for a historical perspective. For my master’s research I was interested in broad-scale patterns in preservation of vertebrate skeletal material in shelf sediments. Most shelf deposits don’t preserve bones and teeth in numerous settings, and as such most prior taphonomic studies focused on single bonebed, skeleton, or lithofacies. The Purisima Formation in California is unique as it preserves abundant marine vertebrates in outer, middle, and inner shelf settings, permitting study of across-shelf changes in preservation. I collected data from over a thousand specimens (many of which I excavated myself) including abrasion, articulation, phosphatization, fragmentation, and polish. As it turns out, the intensity and style of preservation changed remarkably between environments, and was more or less correlated with inferred depositional energy; higher energy settings nearer shore had lower rates of articulation but higher rates of fragmentation, abrasion, phosphatization, and polish. Even more extreme taphonomic damage characterized skeletal material from bonebeds, which were formed largely by a combination of seafloor erosion during transgressive episodes followed by periods of nondeposition. Preservation also varied between taxonomic groups and tissue types – regular bones were somewhat average, with teeth more durable, and calcified cartilage less so. Bird bones, perhaps owing to their lower mass, were less taphonomically modified; cetaceans had some of the highest rates of damage, particularly fragmentation (perhaps according to osteoporosis). These differences indicate a potential for bias regarding differential preservation, and suggests a significant taphonomic overprint.

This paper was a long time in the making, and had its roots in a comprehensive phylogenetic analysis of pinnipeds that Morgan and I started putting together back in 2008 or 2009. We presented on it in 2010, and quickly realized that the way forward was to break up the analysis by family. Our first stab was the phylogeny presented in the Pelagiarctos article, where we added a few new mandibulodental characters for walrus phylogeny. Pinnipeds are certainly challenging to work on; historically there has been a bit of debate about which characters to use, and there has also been some rather shocking choices in character selection, and some parts of the pinniped skeleton have been totally neglected. Anyone who has read this blog before knows that pinniped phylogeny is fairly contentious, and perhaps not as politely discussed as various disagreements in cetacean phylogeny, for example. Disagreements over whether pinnipeds are monophyletic still exist despite diphyletic hypotheses being thoroughly discarded by molecular studies decades ago, and disagreements over the sister taxon of walruses and desmatophocids (among others) still rage. While we will be revisiting odobenid phylogeny with a far larger and more comprehensive analysis in the future (with spectacular new fossils… yay!), our second foray on our long-term pinniped cladistics ‘campaign’ focused on my other favorite group of pinnipeds, the otariids – fur seals and sea lions. Unknowns in otariid phylogeny include 1) are fur seals and sea lions reciprocally monophyletic (e.g. Arctocephalinae and Otariinae), and 2) when did otariids make it to the southern hemisphere? We set out to clarify these questions using morphological and molecular data. Fur seal monophyly had been proposed by an earlier study using fewer characters, but has been challenged by most (if not all) molecular studies which suggest that arctocephalines are paraphyletic, if not polyphyletic. Our results confirm (for the first time using morphological data) fur seal paraphyly, but in general there is a fair amount of variation in the trees from different types of analyses and search protocols. This is not quite our fault, but likely relates to problems with fur seals and sea lions: they all look the same, resulting in relatively few characters distinguishing between different otariids. Furthermore, there is extreme variation amongst these characters, further obscuring phylogenetic signal. Regardless, we did the best we could, and produced the best analysis to date (and you can see all of our data on morphobank!). Using updated dates of fossils from the southern hemisphere, our results permit reconstruction of the southern hemisphere dispersal of otariids to the latest Miocene and along the east Pacific margin, during a period of unusually cool equatorial surface temperatures in the eastern Pacific. What do we need to further the study of otariid evolution? More fossils and more characters, of course! Anyway, next up in the pinniped phylo campaign is a serious reappraisal of odobenid phylogeny (my focus), and an even more critical reassessment of phocid phylogeny (Morgan’s focus).

Modern pinnipeds consume a variety of prey items, including fish, krill, bivalves, cephalopods, and even birds and occasionally other pinnipeds. Different feeding behaviors are utilized including regular raptorial feeding (often accompanied with suction), suction feeding, filter feeding, and macrophagy (e.g. ripping large prey items into smaller pieces for swallowing). Such a diversity in feeding was likely present in fossil pinnipeds. How do we tell how extinct pinnipeds ate? We can take a stab from the “ass end”, so to speak – see if there preserved gut contents (or, coprolites) as often are for marine mammals. Well, there are only two published cases of pinnipeds with gut contents: one is the phocine Kawas from Argentina, and another is a Pliocene phocid from Waihi Beach in Taranaki, New Zealand, in a private collection but “published” in an abstract by Joe McKee. Kawas tells us that (unsurprisingly) it ate fish (in other news, the sky is blue). Rather than using the end-product, we can also attack it from the business end – seeing if differences in feeding morphology can be used to predict feeding ecology, by analogy with observed feeding adaptations in modern pinnipeds. To tackle questions like this, Morgan Churchill and colleagues used discriminant function analysis of tooth spacing measurements for modern and fossil pinnipeds. Earlier studies have found that only a rough correlation between feeding morphology and ecology exists, as most pinnipeds are dietary generalists – a finding supported by this study, which is the most thoroughly quantitative on the subject to date. Pinnipeds that either filtered prey items from the water column or ripped prey to pieces before swallowing generally have larger teeth and smaller gaps between them than pinnipeds which just swallow prey items whole. As opposed to earlier studies which identified four different strategies – pierce, suction, filter, and grip-and-tear feeders, this study found support for three basic groups: sieve feeders like Lobodon and Hydrurga, large-toothed raptorial feeders, and small-toothed raptorial feeders. Mirounga belongs to the latter category, likely reflecting tooth reduction and specialization for suction feeding – although critically, the bearded seal (Erignathus) has been observed suction feeding and plotted in this study with other phocine seals, and further, the crabeater and leopard seals (Lobodon and Hydrurga) were not predicted to be suction feeders – effectively, this not only attests to the limited morphological adaptations for different feeding styles but also highlights the behavioral plasticity of pinnipeds. Most critically, and why this study is on this list as it is otherwise largely neontological in scope – is that it reconstructed the likely feeding ecology of the fossil pinnipeds Enaliarctos and Desmatophoca. Both extinct taxa plotted close to modern otariids, suggesting that they were piscivores that used pierce/raptorial feeding behavior. Part of Morgan’s original question was to figure out whether or not Enaliarctos (the earliest known pinniped) – which primitively retains carnassial teeth – was able to feed on fish and swallow them whole, or whether it needed to return to shore to chew its food like its terrestrial ancestors.

Cope's rule (advanced by none other than Edward Drinker Cope) is the hypothesis that body size tends to increase through time within verebrate lineages (I'm not exactly sure if it was proposed to apply to invertebrates). This idea sought to explain trends toward gigantism, followed by extinction of a lineage, niche replacement, and trend toward gigantism in the replacing taxon. Cope was inevitaby inspired by gigantic fossils of sauropod dinosaurs, but also worked a fair amount on fossil baleen whales - which are quite a bit larger today than the majority of most fossils. Cetaceans are the obvious group to study given the amazing diversity of sizes and gigantic modern sizes - but their fossil record is a bit problematic to work with in some ways. My buddy Morgan Churchill published another recent article (not included here because it is neontologic rather than paleontologic) proposing equations to predic body size from craniomandibular measurements as part of his Ph.D. - and these were applied to fossil pinnipeds in order to examine trends in body size through time. Cope's rule - defined as consistent trends towards larger size within a lineage - is present at a very rough level; the earliest pinnipeds are generally quite small, and modern pinnipeds can be quite large. However, rather than showing a universal increase in body size amongst various lineages, body size tends to diversify through time - in other words, maximum body size increases (cf. walrus, elephant seals), but minimum body size does not (cf. baikal seals, townsend fur seal) - which appears to reflect passive radiation into different niches.


Note: I’m bundling these two papers together since they tell the same story in differing levels of detail. Several species of sea lions are alive today, including the northern hemisphere California (Zalophus californianus) and Steller’s (Eumetopias jubatus) sea lions, and southern hemisphere species such as the South American (Otaria byronia), Australian (Neophoca cinerea), and New Zealand sea lion (Phocarctos hookeri). I’ve seen many of the latter while living down under, and they are mean: Zalophus back home in California aren’t very bold and often will return to the water if you approach them, but I’ve seen bull NZ sea lions charge idiot tourists with cameras. Although easy to see along nearby places like the Otago Peninsula, they don’t actually breed on the mainland – something that has puzzled biologists for quite some time. They breed at offshore islands, like the Subantarctic islands for example, and simply haul out on the mainland – but never stay very long. This is puzzling, as Holocene subfossils of juveniles reported by paleornithologist Trevor Worthy demonstrate that in prehistory they bred onshore. These new studies by Collins et al. report that subfossil Phocarctos from mainland New Zealand are genetically distinct from the modern population using analysis of ancient DNA, and bred on the mainland until becoming extinct. It’s unclear what this means as far as taxonomy is concerned – if anything, it might mean it’s distinct only at the subspecies level. Prior to Māori arrival, specimens that are molecularly related to the modern form are found only in sites on the Subantarctic islands. These studies demonstrate the extirpation of an entire population of mainland breeding sea lions (fossils of which indicate it inhabited all of the mainland aside from the northwestern coast of the South Island) shortly after Māori arrival (indeed, many of the specimens sampled are from middens) and the area reinhabited by the subantarctic population of Phocarctos, but retaining the ancestral subantarctic breeding behavior. Interestingly, this parallels the mainland extinction of the Waitaha penguin; within a century the areas formerly occupied by the Waitaha penguin (Megadyptes waitaha) were recolonized by the extant population of closely related Yellow-eyed penguins (Megadyptes antipodes). As an aside, the second author on these papers is friend and fellow Dunedinite Nic Rawlence.

Modern porpoises (Phocoenidae) are relatively small bodied dolphins that generally lack “beaks” – they inhabit much of the North and South Pacific, and North and South Atlantic. Extinct porpoises have similar distribution, but are most abundantly known from the North Pacific – and a North Pacific origin for the clade has been implicated. A late Miocene dispersal to the eastern south Pacific is evident thanks to fossils like Australithax from Peru; a single Pliocene periotic from offshore New Zealand indicates the timing of dispersal to the western South Pacific. The North Atlantic is another story; phocoenids are totally absent from Pliocene rocks in the eastern USA. In 2008, an archaic phocoenid, Septemtriocetus, was reported from the “middle” Pliocene of Belgium. By this point in time the Panama seaway was already closed to trans-oceanic dispersal, but the Bering strait had opened – indicating that the ancestor of Septemtriocetus likely invaded the North Atlantic via the Arctic Ocean (which was not yet glaciated). A new phocoenid, named Brabocetus gigasei, is reported in this new study from the lower Pliocene Kattendijk Formation of Belgium – somewhat older than Septemtriocetus, indicating an older dispersal to the Atlantic. Brabocetus has an asymmetrical skull, and a well-preserved braincase – effectively resembling a modern Phocoena but with an asymmetrical facial region and a vertex – similar in many regards to Septemtriocetus, except having a wider antorbital notch. The occurrence of these two porpoises in the North Sea Pliocene possibly suggests two dispersals to the North Atlantic or perhaps in situ speciation; regardless, the authors indicate that the harbor porpoise likely represents an independent, late Pleistocene dispersal from the North Pacific.

In 1741, German naturalist Georg Willhem Steller set foot on Alaska for a few hours and became the first European to visit Alaska. This was but a brief stop for the Second Kamchatka Expedition, later known as the Great Northern Expedition - one of the largest naval expeditions, commissioned by Emperor Peter the Great of the Russian Empire to explore the eastern reaches of Siberia (recently incorporated into the empire). On the return from the eastern North Pacific, the crew succumbed to bouts of scurvy, leaving only a dozen able bodied sailors on board, and in fall 1741 the remaining ship, the St. Peter, was shipwrecked on a small island in the western North Pacific. The leader of the expedition, Vitus Bering, died on December 8, and the crew was largely marooned on the island - later named Bering Island. Steller soon discovered bizarre, gigantic marine mammals around the island, which resembled a manatee but had cetacean-like caudal flukes and much larger - up to 8-9 meters in length - and fed upon kelp. It would later be named Steller's Sea Cow (Hydrodamalis gigas) in his honor. It was very buoyant, and part of the sea cow's back was always exposed above the water line, and was otherwise a very slow swimmer unafraid of humans. They could be killed simply by using grappling hooks and dragging them ashore - and worse, they appeared to have strong individual bonds, and would attempt to come to the aid of those being killed - which gave hunters an easy next target. The crew eventually built a smaller boat some time after the wreck of the St. Peter was destroyed by foul weather - and they successfully sailed back to Russian settlements in Kamchatka. News of gigantic, easily killed marine mammals spread, and over the next few decades many crews stopped by Bering Island to take sea cows as their fat was excellent for burning (apparently smokeless) and served well as a butter substitute. By 1768, less than 30 years later, all that remained were bones. Conventional thought in sirenian biology considers this to be the last holdout of the sea cow in the Pacific, although Pleistocene and Pliocene fossils indicate that Hydrodamalis formerly lived as far south as Japan and Baja California. The new study by Crerar et al. reports new bones of Hydrodamalis from St. Lawrence Island, also in the Bering sea but further north and east - which date to the Medieval period (800-1150 AD). The samples were actually first discovered by ivory dealers who were using the dense bone to carve knife handles. Differences in Nitrogen stable isotopes indicate a slight difference in feeding ecology than the Bering Island population (which makes sense given the geographic separation from Bering island). They hypothesize that perhaps climatic deterioration combined with aboriginal hunting by the newly arrived Inuit led to a second, earlier extirpation of a second historic population of Hydrodamalis.

Pinnipeds are sexually dimorphic - males are larger than females and are more robust to boot. Most modern pinnipeds are sexually dimorphic, although it is minimal in most true seals (Phocidae). Extreme sexual dimorphism characterizes the walrus, all otariids, elephant seals, and the gray seal. Sexual dimorphism is most extreme in polygynous pinnipeds - those that breed en masse at rookeries, compete for space, and tend to breed in harems where competition for mates is strong. How can we study the evolution of sexual dimorphism? One recent study used only data from modern species only and promoted an ultimately flawed hypothesis that the earliest pinnipeds were not only minimally sexually dimorphic but also primitively ice-breeders. This could have been easily checked using a couple brief readings of the pinniped paleontological record - none of the earliest known pinnipeds, enaliarctines, are known from polar regions, and enaliarctines like Pteronarctos goedertae are already thought to have been sexually dimorphic. The distribution of sexual dimorphism amongst all three families does suggest that it is primitive for the group - and many extinct pinnipeds on the stem of extant families (Thalassoleon, Otariidae; Neotherium, Imagotaria, Dusignathus, Valenictus, Odobenidae; not sure about extinct phocids) also demonstrate widespread sexual dimorphism in the past. As with all evolutionary questions, if a question can be answered with fossils - our only source of objective information about the history of biological structures - it should. It's actually quite easy, since sexual dimorphism is reflected in the skeleton - and this new paper by Thomas Cullen and others does just that. They use a morphometric dataset including a few extant phocids, extant otariids, the walrus, the extinct pinniped Desmatophoca oregonensis, and the enaliarctine Enaliarctos emlongi. Two skulls of E. emlongi are known - the well-preserved male holotype, and a smaller, crushed skull they reasonably interpret as a female. Their morphometric analysis (Procrustes principal components analysis) quantifies the descriptive statements about pinniped dimorphism, and makes several new morphological observations on dimorphic features; females have been known to have smaller nuchal and sagittal crests for a long time, but Cullen et al. also found that they have proportionally narrower palates and rostra, and have a less robustly constructed squamosal (specifically the mastoid region and the paroccipital process of the exoccipital). Using this data they executed an ancestral character state reconstruction, which reconstructs (accurately in my opinion) sexual dimorphism as being the primitive condition for all pinnipeds. However, there are a couple issues with this study. First, it is not the first study to propose that enaliarctines were sexually dimorphic: Annalisa Berta published evidence for this in Pteronarctos goedertae back in 1994, though not acknowledged by Cullen et al. Secondly, Cullen et al. only compared the supposed female skull with Enaliarctos emlongi, E. barnesi, and E. tedfordi - but not with Pteronarctos (which occurs rather low in the Newport sequence) and Enaliarctos mitchelli, which is a tiny narrow-snouted taxon also reported from the Nye Mudstone by Berta in 1991. This specimen could easily represent another E. mitchelli given its tiny size, so it is problematic that no comparison was made.

It’s (mostly…) been a good year for marine vertebrate taphonomy (there are several new marine reptile and whale fall papers from 2014 that don’t really fit on this list). A few papers over the last few years have detailed some taphonomic advances on baleen whale skeletons from the Pliocene of Italy. A new paper by Silvia Danise and Stefano Dominici takes a quantitative approach towards taphonomic aspects of 25 mysticete skeletons from shallow marine deposits, mostly from the Pliocene (but one Pleistocene and one Miocene specimen as well). As discussed above (see Boessenecker et al. 2014), it’s difficult to study modern taphonomic processes on the continental shelf – currents and rapidly shifting sediment make burial or transport of carcasses and skeletal elements a certainty, aside from being shallow enough to permit carcass refloating. Do differences in sediment, fauna, and depositional processes translate into differences in preservation between the deep sea (where we have plenty of experimental data, but essentially very few paleontological examples of true deep sea occurrences) and the continental shelf (where the sheer majority of marine vertebrate fossils were deposited, but have virtually no actualistic data)? We should expect differences. Danise and Dominici report a number of interesting and some previously predicted patterns – one such predicted example is the lower rate of skeletal articulation in sandstones as opposed to mudrocks (I reported the same result). Several headless skeletons were interpreted to be carcasses that had refloated and disarticulated while floating prior to final sinking, as observed by aktuopalaontologist Willhelm Schafer in the 1930’s (from a rowboat in the North Sea). They report a high number of shark teeth associated with skeletons, including articulated skeletons and (correctly from my perspective) argue that shark scavenging is ineffective at disarticulating large carcasses (also confirmed by whale fall experiments) but also making the case that teeth are only common around skeletons, reflecting scavenging. It’s also possible that shark teeth are just as common in background sediment, and without looking at the background sediment, it’s technically not a great argument to make (although this is me being nitpicky). They also mention bone degradation by microorganisms or Osedax, but make no mention of the actual traces – which from the perspective of a reader, does not really eliminate the possibility that the bones are just poorly preserved or were damaged during collection or preparation. Several skeletons have associated invertebrates, but only one has any specialist chemosynthetic (e.g. whale fall) mollusks; the rest are associated with scavenging or filter feeding mollusks. So, my question is this: if only one skeleton is demonstrably associated with a fossil whale fall invertebrate fauna, isn’t only one out of 25 specimens a whale fall? Regardless, this paper demonstrates that there is a far higher variability in preservation on the continental shelf – probably reflecting the variation in lithology, fauna, and depositional processes on the continental shelf. If you have an interest in marine taphonomy, this is a must-read and one of the most important studies in this small subdiscipline in years.

How did archaeocete whales hear? We know that modern odontocetes and mysticetes hear in two different ways: at high and low frequencies (respectively). High frequency hearing is excellent for echolocation, as the sounds travel (relatively) slowly and not very far. Low frequency vocalizations by baleen whales are loud, and travel far and fast – permitting blue whales to communicate across vast stretches (hundreds of kilometers) of ocean (although the classic humpback whale song is what most will think of). Why is hearing frequency so important? One famous baleen whale individual (yes, a single individual recorded underwater since 1989) of unknown species – known to the public as the world’s loneliest whale – vocalizes at a frequency higher than the range of hearing of all other baleen whales, rendering it incapable of being heard by members of its own species (whichever one it belongs to). Sadly, this individual travels the globe by itself with nobody – aside from humans with some microphones – to hear it. But dead whales sing no songs, so how do you determine the hearing capabilities of extinct whales? The hearing frequency range can actually be estimated by looking at a number of features of the inside of the inner ear, including the easiest to explain example – cochlear coiling. The cochlea is a spiral-shaped organ where currents in the fluid inside pass over little hairs and are transmitted through nerves to your brain; the currents in the cochlear fluid come from vibrations of the stapes, incus, and malleus (the ossicular chain: known informally as the stirrup, anvil, and hammer, respectively) which vibrate thanks to amplified sound waves that hit the ear drum. It more or less works in the same general pattern in cetaceans, with a few differences that need not be explained here – rather, the take home message is that a more tightly coiled cochlea reflects low frequency hearing, and a loosely coiled cochlea reflects high frequency hearing. (Note: all of this happens within the periotic bone which paleocetologists obsess over so much). Early studies used serially sectioned archaeocete periotics - serial sectioning is the slow grinding of fossil, with photos or diagrams recorded incrementally to put together a 3D model and cross section; contemporarily the results from serial grinding could be called a “bootleg CT scan”. There’s no excuse for serial grinding anymore, since we have CT technology which (best of all) actually leaves you with a complete specimen afterwards – after serial grinding, you’re left with the data you collected and a pile of bone dust. Serial grinding studies originally suggested that archaeocetes like Zygorhiza were high-frequency adapted. CT scans of Zygorhiza taken by Eric and Rachel indicate that hearing was most similar to modern baleen whales – and that a number of features, including cochlear coiling, are adaptive for low frequency hearing. This is fascinating and perhaps not surprising from an ecomorphologic perspective, as the earliest mysticetes were effectively archaeocetes with broad rostra, whereas contemporaneous Oligocene dolphins like Cotylocara, Simocetus, and Waipatia already had most of the adaptations for high frequency hearing and echolocation present (see Geisler et al., below) – suggesting that whatever the earliest mysticetes were doing from an ecological perspective, it could not have been too much different than basilosaurids. Also interesting is the fact that Zygorhiza appears to have been more sensitive to rapid head movements than baleen whales, suggesting higher agility amongst basilosaurids.

This study stems from Joe El Adli’s undergraduate research at San Diego State University, which he continued to work on after getting a job at the San Diego Natural History Museum. The SDNHM had on loan from UCMP a beautiful little skull of Herpetocetus from the Pliocene San Diego Formation. I was of course interested in this specimen thanks to my own work on Herpetocetus bramblei from the Purisima Formation. The San Diego Fm. sample of Herpetocetus is much larger, and the fossils represent a somewhat different species than that from the Purisima; the San Diego Formation is also slightly younger. Herpetocetus morrowi is a small mysticete, perhaps only 4 meters in total length when alive, with a rather peculiar jaw joint that would have restricted opening of the mouth to about 15-25 degrees or so – quite the opposite from balaenopterid rorquals, where the mouth can open to over 90 degrees. Herpetocetus has an archaic braincase with highly “telescoped” rostral elements, which others have tied to lunge feeding. However, lunge feeding does not appear to have been possible, owing to the restricted gape. Instead, this critter has a number of adaptations for longitudinal twisting of the jaw, in a fashion similar to a gray whale. We make the case that, based upon anatomical evidence, Herpetocetus was also a suction/benthic feeder – all other cetaceans with adaptations that limit oral gape are suction feeders, and longitudinal rotation could be used to laterally scoop up prey laden sediment into the mouth. This interpretation dovetails quite nicely with a prior study of mine that reported a middle Pleistocene extinction of Herpetocetus as well as a study by Nick Pyenson and Dave Lindberg which identified that glacioeustatic fluctuations in sea level would have forced gray whales (Eschrichtius) to maintain alternative feeding strategies (as low sea levels during glacial maxima would have decreased the amount of available shelf space for benthic prey). Because Herpetocetus was incapable of lunge feeding and appears to have been overspecialized for benthic feeding, we hypothesize that increasingly more extreme middle Pleistocene glacioeustatic sea level changes drove this dwarf baleen whale to extinction.

Link here

This new paper is the most detailed in a series of papers by a group of “creation geologists” from Loma Linda University about the taphonomy of fossil whales in the Pisco Formation of Peru. They present a rather large body of data collected from field studies in Peru. Data they report is similar in scope to that reported for marine vertebrates from the Purisima Formation (Boessenecker et al., 2014, above) with the exception of having more detailed sedimentologic descriptions. Fossil whales from the Pisco Formation are frequently articulated and preserved in diatomite, and occasionally preserved with baleen – which is not surprising, preserved baleen from these localities has been known since the 1980’s. Notably and unfortunately, these authors chose not to excavate and collect the fossil cetaceans: they partially excavated them, and then reburied them. In a way this makes the work untestable, with the exception that you *could* go out to the Peruvian desert and dig up another batch of skeletons (on the contrary, every single specimen from Boessenecker et al. currently resides in museum collections so that my results are readily auditable). They conclude based upon taphonomic (rather than sedimentologic) data that the whales must have been buried rapidly, and (rather strangely in my opinion) state that Osedax bioerosion would likely destroy the skeleton if exposed longer than a few weeks; they cite several modern studies to support this assertion, despite none reporting complete Osedax degredation of a skeleton in under one year (and that is only one example; many skeletons at whale falls have persisted for decades). They do acknowledge that the carcasses could have sunk into soupy sediment (the precursor to diatomite is called diatomaceous ooze for a reason), but strangely do not modify their carcass burial rate estimates accordingly. What I must note is that outside the peer-reviewed scientific literature, these authors have published creationist articles that extrapolate these problematic burial rates to the entire Pisco basin, claiming that the basin would have been filled in a number of millennia rather than over a 15+ million year period as demonstrated by real dating methods. They then use this to attack absolute dating methods and claim that radiometric dating does not work. This information is not exactly hidden deeply on the internet, and one need not spend much time on google to find out that none of these authors really believes what they publish in scientific journals. In this context, all sorts of holes start to appear. For example: if one adopts a disingenuous approach, why be forthcoming with all of your data? Perhaps the rate of articulation is exaggerated by the authors only excavating what upon discovery were clearly articulated skeletons. Where are all the dolphins and pinnipeds? And marine birds like penguins? The authors hardly make any mention of non-baleen whales – indicating a bias towards large skeletons. Not all fossil vertebrates in the Pisco are articulated skeletons – what about isolated elements? Reports of such occurrences abound. What about fossil material in existing museum collections? What about bonebeds and layers clearly indicating massive gaps in depositional history? There is no mention, anywhere, of time-rich bonebeds and preservation therein. What may look to the uninitiated as a piece of solid science has a number of problems, and because this publication is being used to push a Young Earth Creationist paradigm outside the published literature, I cannot inherently trust its content.

Marine mammals have previously been reported from upper Pleistocene sediments from upstate New York, Vermont, and Quebec. Anyone who has visited the famous 18th century star-fort Fort Ticonderoga in upstate New York would be rather quizzical if you told them that during the end of the last glaciation, an inland sea occupied that site (and not, for example, a glacier or a forest). After the last glacial maximum, melt water from the Laurentide ice sheet sat in an isostatic depression (one formed by compaction of the earth’s crust due to massive ice loading) and pooled up – forming the Champlain Sea, which was substantially larger than modern Lake Champlain and included what is now Lake Ontario. Marine mammals from Champlain Sea deposits include species now commonly associated with far colder climates of the Arctic, such as belugas (Delphinapterus leucas), some of the only known fossils of narwhal (Monodon monoceros), several cetaceans dubiously identified from vertebrae and ribs (e.g. Phocoena, Balaenoptera), walrus (Odobenus rosmarus), harp seal (Pagophilus groenlandica), bearded seal (Erignathus barbatus) as well as temperate harbor seals (Phoca vitulina). This paper reports a ringed seal (Phoca hispida) from the Champlain Sea, another arctic phocid. Other ringed seals have been reported before, so this is not a surprise, but this new specimen has yielded radiocarbon dates that indicate marine mammals colonized the Champlain Sea almost immediately after it formed.

Oligocene cetaceans are generally rare worldwide, but bridge the gap between modern cetaceans and Eocene archaeocetes. While the archaeocete-mysticete transition is fairly well known, perilously few early diverging odontocetes have been described. Oligocene odontocetes are not special by virtue of being Oligocene in age; Waipatia and many other Oligocene dolphins are members of the Platanistoidea, one of the earliest diverging dolphin groups with extant members. Other Oligocene dolphins have proven problematic; Xenorophus and Agorophius are known from the Oligocene of South Carolina, but the nearly complete skull of Agorophius has been missing for a century, and Xenorophus is known only from a partial rostrum. A new extinct dolphin from this same locality - and closely related to the poorly known Xenorophus - is described in this new paper as Cotylocara macei. Xenorophids are quite poorly known in the published literature, but a number of unpublished specimens has meant that many paleocetologists are familiar with their morphology, despite not being published; fortunately we no longer have to rely upon unpublished "conventional wisdom". This new odontocete is spectacularly preserved, and is held in the collections of the College of Charleston Natural History Museum. Cotylocara has an attenuate and downturned rostrum with at least 12 double and single rooted teeth (meaning that it is slightly polydont), large fossae on either side of the rostrum, a blowhole positioned slightly anterior to the orbits, and an asymmetrical facial region of the skull. It also has a beautifully preserved but very strange periotic, and is one of the only odontocetes I've ever seen with an archaeocete-like superior ridge. The asymmetry of the skull, and phylogenetic position of Cotylocara as one of the earliest diverging odontocetes, indicates that echolocation in odontocetes evolved relatively rapidly at or around the Eocene-Oligocene boundary. Curiously, Cotylocara is also one of the only known odontocetes to have had a longitudinally twisted rostrum such as recently reported for basilosaurid archaeocetes by Julia Fahlke and others.

Late Eocene assemblages of marine mammals are common and geographically widespread; Late Eocene archaeocetes have near-worldwide distribution. However, early and middle Eocene marine mammals are rare outside the Tethyan region (Mediterranean Africa, India). This study reports one of the first marine mammal assemblages from sub-Saharan Africa, collected from commercially mined phosphatic bonebeds of Togo, a small country in between Ghana and Nigeria. Marine mammal remains include fragmentary postcrania of protosirenid and dugongid sirenians, a new genus and species of protocetid archaeocete they name Togocetus traversei, and additional isolated teeth and vertebrae demonstrating the presence of two additional protocetids – one larger and one smaller than Togocetus. Togocetus is represented by a large number of isolated bones and teeth, including skull fragments, a couple of nice mandibular fragments with teeth, and a tympanic bulla. All the material assigned to Togocetus are of approximately similar size, and when multiple elements are present, are similar in morphology. This is admittedly not a very popular manner in which a hypodigm is assembled; ideally, we would refer specimens that clearly overlapped – but this is a very tetrapod-specific approach, and generally not followed by shark paleontologists who must deal with a fossil record which is 99.9% composed of isolated teeth that must be divided into groups. Regardless, hypodigm assembly such as this has its merits, and is totally reasonable as it is a testable hypothesis – and is testable by going out and digging up more fossils.

Beaked whales have all sorts of weird bony structures of the rostrum, facial region, and mandible, including large paired fin-like crests (bottlenose whales, Hyperoodon), extraordinarily dense rostra (Mesoplodon densirostris), and densely ossified mesethmoid (normally cartilage; Cuvier’s beaked whale, Ziphius cavirostris). It’s even weirder when you look at fossil ziphiids – some have little bony horns on the maxillae (e.g. ), a large median ridge or fin of dense bone on the rostrum (Tusciziphius, Caviziphius), an enormous spherical median nodule of bone on the rostrum (Globicetus), and longitudinally swollen premaxillae (Ziphirostrum, Aporotus). Histological study of modern and fossil ziphiids indicate that many of these different facial ossifications arise via different pathways, indicating developmental diversity as well as structural diversity. No single good argument has been made to explain these structures; in modern ziphiids they are of course sexually dimorphic – so explanations relying upon dense bone for buoyancy and acoustics (e.g. facilitating hearing) can be rejected outright, since these structures are absent in females. Rather, they suggest a behavioral role. Pavel Gol’din put more thought into this, and suggested that these are auditory “display” structures. These were never considered for display since they are internal to soft tissue and would not be visible externally. However, odontocetes can echolocate and experimental data indicates that dolphins can visualize 3D information from echolocation. Putting two and two together, Gol’din hypothesizes that these disparate skull structures would be easily detected from echolocation – and not only species, but sex would be readily apparent from acoustic visualization (as ziphiid skulls have robust sexual dimorphism). This in my opinion is a totally fascinating hypothesis, and one that makes surprisingly good sense.

Archaeocete whales are most famous for retaining functional hindlimbs, and many early archaeocetes like Pakicetus and Ambulocetus could walk or run on land – and their appendicular skeletons looked not too dissimilar from other artiodactyls. Basilosaurids have reduced hindlimbs and a pelvic girdle decoupled from the vertebral column, and are thought to have facilitated copulation similar to the vestigial hindlimbs of boa constrictors. Aside from the singular fact that modern cetaceans retain a vestigial pelvis (the only reported function of which is to anchor muscles that move the penis; not sure about function in females), most paleocetologists who work on Neoceti are unashamedly cranial in focus, something I began to lament when I started describing postcrania of NZ eomysticetids. The focus on craniomandibular material has led many neocete workers ignore or give poor descriptions of postcrania of fossil neocetes. So, what happened in between having an external hindlimb in the Eocene and a tiny surfboard-shaped pair of pelvic bones in modern cetaceans? This new paper (also by Pavel Gol’din) describes several well-preserved innominata (pelves) and partial femora and a tibia of mid-late Miocene cetotheriid mysticetes from Paratethys (southeastern Ukraine). These have quite a bit more morphology than the simplified “pelvic surfboard” of modern cetaceans and permit identification (and clarification) of homologous features on the cetacean pelvis. These cetotheriid pelves are three-pronged, and lack a socket for the femur or an obturator foramen (large hole between the ischium and pubis), and have a posteroventrally descending pubis. Basilosaurid pelves have an obturator foramen and a socket for the femur (acetabulum), but the orientation of the innominate is unknown and different interpretations have been published by Kellogg in the 1930’s and Phil Gingerich in the 1990’s. Pavel’s study suggests that Kellogg’s early reconstruction – with the hip joint at the posterior, rather than anterodorsal end of the pelvis – is likely correct. This also suggests that the left and right innominates may not have been medially connected at a pelvic symphysis as hypothesized by Gingerich, but may have been floating in soft tissue as in all modern cetaceans (which is another point of debate in current archaeocete literature; see recent publications by Hans Thewissen for more).

This paper is a followup to Gol'din and Zvonok (2013) who named a new genus and species of basilosaurid archaeocete whale from the Eocene of Ukraine, Basilotritus uheni. The new genus also included B. wardii, a species formerly attributed to the protocetid Eocetus. Basilotritus has distinctive pachyosteosclerotic vertebrae with abundant external foramina. This new specimen was collected as an associated skeleton from the upper part of a glauconitic hiatal deposit rich in shark and fish remains, and includes mandible fragments, numerous teeth, vertebrae, ribs, a well preserved sternal skeleton, and fragments of the scapula and a single phalanx. The teeth are similar to other basilosaurids but differ in having secondary denticles on the accessory denticles (e.g. the denticles on the postcanine teeth are almost serrated) - a unique feature amongst archaeocetes - but primitively retain inflated tooth roots like protocetids. As in other archaeocetes, the sternum is multielement - unlike the single-element sternum of mysticetes, for example. In general, the dental and skeletal anatomy of this specimen confirms the transitional morphology of Basilotritus with a mix of protocetid and derived basilosaurid features and cladistic analysis demonstrates its position as an early diverging basilosaurid. However, the cladistic analysis uniquely shows a Neoceti + Ocucajea (another archaic basilosaurid, from Peru) clade as the earliest diverging lineage within a paraphyletic basilosauridae, which is surprising given that Dorudon is typically identified as the sister taxon of Neoceti (albeit, perhaps erroneously). Better preserved fossils of basilosaurids, and analyses including a wider range of archaeocetes and neocetes (e.g. not focusing on one with only a couple taxa from the other) are really needed to further investigate the archaeocete-neocete transition. Lastly, the teeth show strongly developed wear facets similar to the protocetid Babiacetus, and perhaps indicative of predation on sharks as reported for modern killer whales. Gol'din et al. further mention the abundance of shark teeth in the same horizon as lending support to this, although since both occur within greensand this is likely depositionally controlled and simply taphonomic coincidence rather than anything paleoecologically meaningful.

There has been an explosion of new ideas and species of “cetotheres” over the past decade, and it seems to be accelerating. This started with a couple of papers in the mid-2000’s by Virginie Bouetel who for the first time cladistically demonstrated that a subgroup of cetotheres (as formerly defined, a totally paraphyletic wastebasket group) actually do share a number of features that unite them into a clade now formally (and widely) recognized as the Cetotheriidae sensu stricto (or, these days, just Cetotheriidae). One fascinating development is the reinvigoration of study of paratethyan cetotheres. Paratethys was an enormous foreland basin that was occupied by a continuous inland sea that stretched from Austria to the Aral Sea in Kazakhstan; the Black Sea, Caspian Sea, and Aral Sea are Holocene remnants of a formerly much larger basin, and fossiliferous deposits are now uplifted and exposed around their margins and in between. Brandtocetus is known from a couple of partial skulls collected from upper Miocene marine rocks on the Crimean peninsula (collected prior to the Ukrainian civil war, of course). The skull of Brandtocetus is similar in many regards to Cetotherium, and lacks the twisted temporomandibular joint of Herpetocetus; the braincase is somewhat wider, and has sinuous nuchal crests. Most importantly, it has well-preserved earbones that are removed from the skull, unique amongst described paratethyan cetotheres; in most other specimens they are still embedded in matrix. This new taxon expands our knowledge of paratethyan cetaceans, and I was quick to code this new genus for my own thesis cladistics. Note: it’s been another damn good year for Pavel Gol’din!

True seals (Phocidae) are the most diverse group of pinnipeds, and have an extensive fossil record in the North Atlantic, Mediterranean, and Paratethyan region; however, the record is less robust in the southern hemisphere and does not include many well-preserved skulls, in contrast to other groups of pinnipeds in the fossil record. One key phocid is the extinct Pliocene taxon Homiphoca capensis, known from the early Pliocene Langebaanweg locality of South Africa. Most of the material is isolated, but only one seal appears to be present. Additionally, Homiphoca has been identified from the early Pliocene Yorktown Formation of North Carolina, indicating that it had an antitropical distribution. The skull of Homiphoca is generalized and quite similar to the Peruvian phocid Piscophoca. Like most southern hemisphere phocids, it is a monachine; monk, elephant, and all Antarctic phocids are monachines – while the Antarctic seals (crabeater, leopard, ross, and weddell) form the tribe Lobodontini. Homiphoca has been generally considered to be closely related to other extinct southern hemisphere monachines like Piscophoca, Hadrokirus, and Acrophoca; however, very few morphology based cladistic analyses of phocids exist. One of the first good analyses was published by Eli Amson and Christian de Muizon last year for Hadrokirus – which recovered Homiphoca, Piscophoca, and Hadrokirus together within a clade that in turn was sister to the Antarctic lobodontines and Acrophoca. This new study reports on the cranial morphology and phylogeny of Homiphoca, using a sample of 7 partial and complete skulls, figured beautifully in the new paper. Each is coded as a separate OTU in the analysis to see if Homiphoca capensis truly is a single species. Monophyly of Langebaanweg seals is confirmed, whereas an earlier morphometric study by Govender suggested the presence of at least two different species. These new results suggest that any future splitting of the Langebaanweg Homiphoca may not be so well-founded, given the poor separation of Homiphoca specimens in the analysis.The new results, unsurprisingly, place Homiphoca within the lobodontine clade – but surprisingly - as the sister taxon of the Ross seal – Ommatophoca rossi – possibly the weirdest looking of all phocids (internally, anyway – their skulls are very alien-looking). Piscophoca is placed as sister to the Ommatophoca-Homiphoca clade, with the other lobodontines and Acrophoca falling just outside, or switching places with Piscophoca depending upon the analysis. The Pliocene Langebaanweg assemblage is notable for the abundance of Homiphoca and complete absence of otariids – today, the cape fur seal Arctocephalus pusillus is the only local pinniped with Antarctic lobodontines occasionally straying into south African waters – suggesting faunal changes in the Plio-Pleistocene. Govender hypothesizes that ancestral lobodontines were primarily accustomed to hauling out on sandy beaches rather than rocky shores (as modern lobodontines generally only haul out on smooth pack ice rather than rocky shores), and dispersed to south Africa via island haul outs; Plio-Pleistocene sea level transgressions resulted in the proliferation of rocky shores, which otariids were able to colonize more easily.

Three extant (Boto, Susu, Franciscana) dolphins and one recently extinct river dolphin (Baiji) are known as river dolphins because they inhabit rivers (the Francisca actually is marine and inhabits estuaries, but is anatomically similar and thus grouped with river dolphins anyway). These dolphins apparently all belong to lineages that independently “adapted” to freshwater environments, although in most cases it’s unclear what anatomical adaptations are really needed. Some river dolphins have smaller eyes, reflecting a stronger use of echolocation and lack of emphasis on vision in dark, sediment-laden water. One unique adaptation – thought to permit extra maneuverability in rivers in the Amazon basin – is the “double” shoulder joint of the Amazon river dolphin, Inia geoffrensis. Aside from the standard humerus-scapula joint, a second joint is present where the proximal humerus also articulates with the sternum – which is much wider in Inia than other odontocetes. Gutstein et al. report an isolated iniid humerus from upper Miocene nonmarine rocks of Argentina, which they tentatively identify as belonging to an ischyorhynchine iniid, which has a proximal end similar to modern Inia. This specimen suggests that the double shoulder joint adaptation has been around for at least 6 million years or so.

This short paper reports a single vertebra they identify as the desmostylian Paleoparadoxia from lower Miocene rocks of Hokkaido. Paleoparadoxia is abundantly known from Miocene rocks of Japan and California, and late Oligocene desmostylians are known from both areas (and Alaska) but include earlier taxa such as Cornwallius, Ashoroa, and Behemotops. Paleoparadoxia has recently been split up into three genera – Archaeoparadoxia from the earliest Miocene of Mendocino County, California, and Neoparadoxia from the late Miocene of Orange County, California; all California specimens were assigned to these two new genera, with the genus restricted to Paleoparadoxia tabatai. In this case, the use of the genus Paleoparadoxia in this paper is now equivalent of Paleoparadoxiinae of Barnes (2013) – who split the genus up. Hasegawa et al. do cite Barnes (2013), but because I cannot read Japanese, I’m not sure what stance they take on the splitting. Regardless, this new specimen demonstrates continuous inhabitation of the western North Pacific by paleoparadoxiines through the early Miocene, and is one of the first records from Japan that is temporally equivalent with Archaeoparadoxia weltoni from California.

Kazár, E., and Hampe, O. 2014. A new species of Kentriodon (Mammalia, Odontoceti, Delphinoidea) from the middle/late Miocene of Groß Pampau (Schleswig-Holstein, North Germany). Journal of Vertebrate Paleontology 34:1216-1230.

This paper reports a fragmentary new fossil named as a new species of dolphin in the genus Kentriodon. Kentriodon is known from the middle Miocene of the Atlantic coastal plain in the eastern US, Sharktooth Hill in the western US, and Japan. This new species, Kentriodon hoepfneri, is known from a fragmentary skull with a well-preserved periotic and partial bulla, mandible fragment with teeth, and a nearly complete vertebral column. Normally I’d be inclined not to treat a specimen like this as name-able, but the periotic is certainly an exception and importantly preserves some differences with other Kentriodon, notably a small, oval-shaped facet for the attachment of the bulla. The Kentriodontidae is another example of a cetacean family where monophyly is assumed rather than demonstrated, as correctly noted by Kazár and Hampe; they note that some of the features identified as supporting this family may be symplesiomorphies (rather, features that are primitive and cannot really be used to diagnose a clade). Interestingly, the completeness of the vertebral column permits functional comparisons with modern odontocetes, and in terms of centrum length profile along the column the proportions are similar to modern beluga (Delphinapterus) and some beaked whales (Ziphius, Mesoplodon). The neck and thoracic column are relatively flexible, while the caudal region is rather stiff; this suggests a skeleton less adapted to rapid, sustained swimming where the vertebrae have similarly stiff articulations throughout the column. Lastly, isolated fossil periotics similar to Kentriodon had previously been reported from eastern Europe, but this new record robustly confirms the presence of Kentriodon in the eastern Atlantic, suggestive of a worldwide distribution during the Miocene.

This study is in Japanese (which I am unable to read) but has an English abstract and excellent figures. This paper reports a late Miocene baleen whale skeleton from central Honshu, Japan. The skeleton is rather fragmentary and includes a very partial skull that is perhaps unusually preserved, well-preserved earbones, and much of a postcranial skeleton although the bones themselves are rather "chewed up". Because of this incompleteness, the authors are not able to confidently identify it to the genus level, and tentatively identify it as a "cetothere" sensu lato (aka stem Thalassotheria, sensu Bisconti et al. 2013). The periotic and tympanic bulla, in my opinion, share many features with archaic balaenopterids like "Megaptera" miocaena and "Balaenoptera" ryani (both from the upper Miocene Monterey Formation of California).

This study reports one of the oldest old world records of a true seal (family Phocidae), a mandible with a few cheek teeth which Koretsky and Domning erected the name Afrophoca libyca upon. The holotype was collected quite recently (2010) by renowned sirenian expert Daryl Domning from the lower-middle Miocene Marada Formation in the Libyan desert. Based on a genial tuberosity that extends to the third premolar, they refer the specimen to the Monachinae. The mandible otherwise has relatively large but simply ornamented cheek teeth, and reflects a seal similar in size to a modern monk seal. The locality is 14-19 Ma in age, indeed making this record somewhat older than most Paratethyan records of seals, and possibly the oldest known record of any crown pinniped (although Desmatophoca brachycephala from Washington, USA might actually hold that distinction). Koretsky and Domning point out that this discovery supports the hypothesis that true seals first arose in the Oligocene in Paratethys, but did not go so far as to claim that it supports the mostly passé idea of pinniped diphyly. As an aside, two interesting and decidedly intentional ‘categorizations’ are worthy of note in the comparisons section: 1) comparisons with the otter-like putative stem pinniped Puijila are placed in the “comparisons with Mustelidae” section, and subsequently in that same section the following statement regarding the dubiously named “pinniped” Praephoca bellunensis (named by C. Diedrich in 2011): “The only known specimen, the proximal part of a femur, is so damaged that it is inadequate for its identification as a pinniped”. Perhaps ironically the same argument could be applied to similarly preserved partial femora from the Oligocene South Carolina described by Koretsky and Sanders in 2002.

True seals – also known as earless seals – have an extensive fossil record in the Atlantic, Paratethys, Mediterranean, and Australasia (and a fragmentary Pacific record limited to the late Pleistocene). However, unlike the record of otariids (sea lions) and odobenids (walruses), the phocid fossil record is depauperate with respect to skulls, and as such the majority of extinct phocid taxonomy is based upon mandibles and “cardinal” fore- and hindlimb elements of the postcranial skeleton such as the humerus and femur. An issue with this is that there is sort of a mismatch in the standards accepted for “otarioid” studies and studies of phocids. For example, although a couple of walruses have been named based upon postcrania (Pliopedia pacifica and Valenictus imperialensis are the only examples), skulls and mandibles are generally preferred. This new paper by Irina Koretsky and colleagues reports three new species of phocids from the upper Miocene Gram Formation of Denmark based upon fragmentary postcranial elements. The first, Pontophoca jutlandica, is based upon a partial femur that bears enough similarities to link it with the middle Miocene paratethyan phocid Pontophoca sarmatica (Moldavia) – supposedly a monachine. The next, a phocine, is named as Gryphoca nordica, and is based on a fragment of a proximal humerus – and placed in the same genus as Gryphoca similis, originally named by Van Beneden from the late Miocene of the Netherlands and later reported from the early Pliocene Yorktown Formation of the Lee Creek mine. The last, Platyphoca danica, is also linked with circum-North Atlantic seal – Platyphoca vulgaris – reported from the early Pliocene of Belgium and also from the Yorktown Fm at Lee Creek. This last one is based upon a distal humerus fragment. How distinctive are these postcranial taxa? Unfortunately not much baseline work exists to evaluate the likelihood of recognizing the “realness” of these species or confident referral of additional specimens.

This study is similar in scope to Churchill et al. (see above), and sought to examine morphological features implicated in feeding ecology amongst modern and fossil true seals (Phocidae). The study does present some measurements, but no analysis is attempted and therefore the findings of the study largely derive from descriptions. Several descriptions of mandibular morphology are given, including elephant, hooded, gray, spotted, monk, and crabeater seals, and four extinct seals: Leptophoca, Pliophoca, Devinophoca, and Miophoca. For some reason dentition is avoided, and the study focuses only upon the mandible itself. The study identified that a deep coronoid process and posterior mandible is characteristic of seals that consume larger prey, while a shallow coronoid process is typical of seals that eat smaller prey such as fish and krill. Extinct seals like the mid Miocene phocine Leptophoca lenis (Calvert Formation, Maryland) and the Pliocene monachine Pliophoca etrusca (Italy) have generalized mandibles with features of both phocines and monachines – but the rather shallow coronoid process suggests a diet of mostly small-medium sized prey items. However, since this study is not quantitative in approach and treats prey size in a qualitative manner in particular, it’s unclear how meaningful these distinctions are.

Modern sperm whales are known for their specialization on eating squid, and the giant sperm whale – Physeter macrocephalus – is often depicted grappling with giant squid. There is a hypothesis, though I cannot recall where I read it, that the giant size of Physeter and the giant squid (Architeuthis) and colossal squid (Mesonychoteuthis) is driven by an arms race between predator and prey. Unfortunately, we don’t really have much of a fossil record of soft bodied squid (and none to speak of for these giants), but the fossil record of sperm whales is robust. Physeter-like (e.g. subfamily Physeterinae, if you will) teuthivorous sperm whales are known from the middle Miocene and include critters like Aulophyester and Orycterocetus – neither of which would have been gigantic, but perhaps the size of a killer whale. Like modern sperm whales, neither possessed upper teeth, and all the lower teeth lacked enamel. In contrast, most other Miocene physeteroids were unlike these derived taxa, and many possessed large teeth with thick enamel crowns set into robust rostra and mandibles and had skulls with much larger attachment areas for jaw closing muscles. A ‘killer sperm whale’ habitus was proposed for late Miocene sperm whales like Zygophyseter and Acrophyseter, but a macrophagous lifestyle was not completely obvious until the terrifying monster Livyatan melvillei was discovered in 2010 – a sperm whale with a 3 meter long killer whale like skull (in terms of proportions and robustness) and teeth the size of 2 liter soda bottles. As a followup to these earlier studies, this paper reports a new skull of the smaller sperm whale Acrophyseter that has a series of pathologies on the side of the upper tooth row. These are a series of bony protrusions (called exostoses) that extend down from the maxilla and sit on the lateral side of the base of the maxillary teeth. These are present posteriorly, and are hypothesized (based on similar formation in better-studied extant mammals like humans) to form as buttresses that support the teeth during loading. Physics dictates that the teeth closest to the jaw joint (the posterior teeth) will undergo the highest loading during biting – and these are exactly the teeth with the exostoses. This study presents additional data supporting macrophagous behavior in Miocene physeteroids.

This paper describes a new squalodelphinid dolphin from Peru – from the same unit as the aforementioned Peruvian Notocetus vanbenedeni (see above – Bianucci et al.), the lower Miocene Chilcatay Formation. Up until now, the Squalodelphinidae included two key taxa with complete skulls: Squalodelphis fabianii from the lower Miocene of Italy, which is known from a complete but otherwise poorly preserved skull with little surface detail and difficult to see sutures, and the better preserved Notocetus vanbenedeni from similarly aged rocks in Argentina. Historically, the monophyly of this group has been assumed rather than demonstrated, and this study included the first cladistic analysis with multiple squalodelphinids included – and robustly confirmed – monophyly of this group (assumptions of monophyly for various “families” remains a bit of a pervasive problem in paleocetology). Regardless, Huaridelphis is known by several well-preserved, virtually complete skulls with well-preserved earbones; this odontocete, like other squalodelphinids, has a relatively elongate and tapering rostrum, single rooted teeth (aside from some waipatiids, these cetaceans are some of the earliest odontocetes to evolve single-rooted dentitions). Subtle differences of the braincase, proportions of various braincase structures, and the periotics differentiate Huaridelphis from Notocetus vanbenedeni, as well as having a higher tooth count.

Stenella kabatensis was first described in 1977 by Horikawa from upper Miocene strata of western Hokkaido, and based on a robust skull that clearly belonged to a delphinid. Subsequent workers doubted attribution to Stenella, which is a small and relatively gracile modern delphinid (Spinner dolphins, Stenella longirostris, are a prime example); Waseda University Ph.D. student Mizuki Murakami set out to redescribe this problematic fossil as part of his dissertation. Previous cladistic analysis of delphinoids by Murakami et al. (2013A,B) suggested that S. kabatensis was much lower on the cladogram; additionally, S. kabatensis is 8-13 Ma in age, much older than the inferred early Pliocene molecular divergence date of the Stenella species complex. As it happens, S. kabatensis lacks a bunch of features found in derived delphinids, and actually shares a few features of the facial region, mandible, and hyoid apparatus with killer whales (Orcinus orca) and the problematic fossil delphinid Hemisyntrachelus cortesii (debate continues as to whether Hemisyntrachelus is distinct from bottlenose dolphins, and whether all the species contained in Hemisyntrachelus should just be transferred to Tursiops instead; Murakami et al.'s results suggest that Hemisyntrachelus is in fact distinctive). Because of these differences in morphology and concomitant different placement on the delphinid tree, Murakami et al. erect the new genus Eodelphis, and note that this dolphin is now the oldest known true oceanic dolphin (Delphinidae). For years an undescribed skull briefly mentioned in an old paper from the Monterey Formation of Orange County, California, was accepted as the oldest known delphinid - but because it was not published, its identification as a delphinid cannot really be formally audited. Murakami et al. further note that all Miocene delphinids have been reported from the Pacific (whilst being absent from richly fossiliferous and densely sampled assemblages from the Atlantic and beyond), strongly suggesting a Pacific center of origin for the clade. However, molecular results suggest a separate origin outside the Pacific - discovery of more fossils is needed to further evaluate the center of origin for Delphinidae. This study actually became the featured article for the May 2014 issue of JVP, and I was honored to make the cover illustration (see image below). Thanks again to my friend and colleague Mizuki for inviting me to contribute!

Despite Eodelphis being a replacement name for "Stenella" kabatensis, the genus name Eodelphis has been occupied for nearly a century for the well-known marsupials Eodelphis cutleri and Eodelphis browni. See, the root "delphis" also applies to opossums - for example, the Virginia Opossum is Didelphis. So, Murakami et al. quickly published this followup note proposing the second new genus name Eodelphinus for the Japanese fossil dolphin.

Modern toothed whales are notable for conspicuous skeletal and facial left-right asymmetry, which is associated with echolocation. Soft tissues in the facial region associated with the nasal passages have been modified to produce all sorts of sounds, notably clicks (used in biosonar) and all sorts of whistles. The prevailing hypothesis is that the right nasal passage and associated musculature has been modified for sound production while the left nasal passage is less modified, and perhaps preferred for respiration. As a result, the maxilla and premaxilla are often wider on the right side than the left, effectively making the midline of the skull shift to the right; the vertex (top of the skull) is similarly shifted to the left hand side. Most Oligocene odontocetes have facial regions that are more or less symmetrical (or, about as symmetrical as terrestrial mammals), but Pliocene and many middle-late Miocene odontocetes have directional asymmetry in the skull. Asymmetry and twisting of the rostrum in archaeocetes (and now, the xenorophid dolphin Cotylocara) is probably not related to echolocation, however. When did asymmetry arise? Did it arise once, or independently several times? Larry Barnes hypothesized that facial soft tissue asymmetry is probably a uniting feature of Odontoceti: all extant odontocetes, with or without cranial asymmetry, bear asymmetrical soft tissues. Thus, Barnes proposed that odontocetes with symmetrical crania likely had asymmetrical soft tissues for echolocation, with cranial asymmetry arising multiple times where soft tissues became modified enough to influence asymmetry in the skull. This new study by Mizuki Murakami reports a partial odontocete braincase representing one of the earliest records of a delphinoid worldwide – a kentriodontid skull from the lower Miocene of Japan. The specimen clearly has asymmetrical nasals, and the midline is drawn onto the specimen by the authors; measurements of the distance of various features from the midline are used to determine the extent of asymmetry. However, the left and right edges of the skull are missing, and it is perhaps not possible to objectively conclude where the midline actually lies. The skull is certifiably asymmetric based upon the nasals – but the placement of the midline is problematic given the incompleteness of the fossil. However, the ancestral character state analysis reconstructed the ancestral delphinoid as having a symmetrical skull – reinforcing the hypothesis of Barnes.

This new study reports a spectacular new assemblage of marine mammals in upper Miocene sandstones from the Atacama desert near Caldera in Chile. These deposits are equivalent in age to some of the younger levels of the Pisco Formation in Peru, and has even produced marine mammals previously known from Peru such as the walrus faced dolphin Odobenocetops and the aquatic sloth Thalassocnus. However, the star of the new paper are not these weirdos but a series of articulated rorqual skeletons (balaenopterids – e.g. the family that includes minke, fin, blue, and humpback whales). Articulated skeletons are concentrated into four horizons. Sedimentologic evidence suggests that this was a sort of supratidal flat with abundant algae within an embayment. The concentration of these skeletons, both vertically, and laterally (many of the skeletons are adjacent) suggests that mass death events at sea resulted in carcasses drifting into this embayment to find their final resting place on the tidal flats – perhaps killed off by an algal bloom. I’m very skeptical of claims of mass death assemblages in the marine vertebrate record since there is such an enormous taphonomic overprint that usually obliterates any genuine ecological signal in the fossil record, but in this case the authors have done a fine job that withstands robust scrutiny.

This paper reports the strangest marine mammal described this year – the bizarre prognathous porpoise Semirostrum ceruttii from the Pliocene San Diego and Purisima Formations of California. Semirostrum is a true porpoise (Phocoenidae) and intermediate in size between bottlenose dolphins and smaller harbor porpoises. It had a mostly symmetrical skull with large premaxillary eminences (a phocoenid feature), and the skull is otherwise unremarkable; all the action is in the lower jaw (mandible). The mandibles are fused at the symphysis (=chin), and the symphysis is dorsoventrally expanded into a toothless paddle shape – the jaw is strongly prognathous, meaning that it extends beyond the tip of the rostrum. This is somewhat reminiscent of the eurhinodelphinids, which were instead extremely retrognathous (=overbite), and had a toothless rod-like rostrum that extended far beyond the tip of the mandible. What could this bizarre structure have functioned as? The symphyseal region has a number of longitudinal canals, which could have housed arteries for a keratinous sheathe or other soft tissue structure, or nerves to permit strong tactile perception (touch). Interestingly, the teeth have strongly developed wear facets but only on the labial (lip/outer) side of the teeth, which rather than developing from wear against other teeth, appears to have been caused by interaction with a foreign substance. We interpreted this toothwear as being caused by sediment abrasion, and that Semirostrum used its strange chin to probe, plow, or stir through sediment looking for soft bodied invertebrates. Notably, several other benthic feeding marine mammals are known from the Pliocene of California, including the toothless walrus Valenictus, one other odobenine walrus, the dwarf whale Herpetocetus morrowi (see above – El Adli et al., 2014), and an unnamed genus of gray whale (Eschrichtiidae). Lastly, I'll note that the reason I was attached to the paper was to include several additional specimens of Semirostrum from the Purisima Formation of northern/central California (Santa Cruz, Half Moon Bay regions) - material collected by myself and dogged amateur paleontologist Stan Jarocki of Watsonville. One skull from the Purisima is somewhat older and perhaps represents a second, older species as it retains obvious facial asymmetry (a primitive trait amongst phocoenids).

Modern odontocetes have all sorts of strange holes in their heads - technically known as sinuses. The most well known sinuses surround the tympanoperiotic, and are thought to acoustically isolate and insulate the inner ear from bone-conducted sound to permit directional hearing underwater. Odontocetes - particularly delphinoids - have a lobe called the pterygoid sinus which extends anteriorly and wraps around and in front of the nasal passages. Phocoenid porpoises have a lobe that even invades between the maxilla and frontal bones in the face; these sinuses are air-filled, causing a huge mismatch in density that is known as acoustic impedence. It's difficult to study these, since these are cavities within a skull - in the old days, skulls would have to be sawed apart to see inside. Nowadays though, CT scanning permits non-destructive 3D imaging of the insides of the skull. My colleague Rachel Racicot has carved out a niche doing this for modern and fossil cetaceans. In addition to the sinuses, CT data can also reconstruct the brain from fossil skulls. This study reports CT findings of the brain and sinus endocasts of the recently described skimmer porpoise Semirostrum ceruttii. Semirostrum does not differ in its cranial nerves or cranial circulatory system from modern porpoises, but bears pterygoid sinus features of both porpoises and delphinids (attesting to its position as a stem phocoenid), and differs from most delphinoids and river dolphins in possessing an unusually robust ossified falx cerebri that penetrates deeply between the cerebral hemispheres. Racicot and Rowe hypothesize that this may be an adaptation towards reducing inertia of soft tissues during rapid movements of the head or perhaps deep diving - whatever the function, it does imply a different sort of behavior for Semirostrum relative to other delphinoids, as we (see Racicot et al., above) hypothesized for this bizarre porpoise.

Kentriodontids are a group of early delphinoids implicated in the ancestry of modern porpoises (Phocoenidae), oceanic dolphins (Delphinidae) and belugas/narhwals (Monodontidae); their fossil record extends back to the early Miocene (see Murakami et al., above) and they more or less have a worldwide distribution (both coasts of North America, Japan, South America, New Zealand, Europe; see Kazár and Hampe, above). The most well preserved and well known kentriodontids are all from the east coast of the United States, principally from the Calvert, Choptank, and St. Marys formations - small dolphins like Kentriodon pernix, Liolithax pappus, slightly larger beasts like Lophocetus calvertensis and Hadrodelphis calvertensis, and the gigantic dolphin Macrokentriodon morani. The monophyly of this "family" has in the past generally been assumed rather than demonstrated, although a few studies like Murakami et al. (2014 - above, on the asymmetric kentriodontid from Hokkaido) have supported monophyly of a subset. In the 1930's, Kellogg named a bunch of kentriodontids based upon isolated periotics from the middle Miocene Sharktooth Hill Bonebed of California (not very good practice) which turned into a bit of a mess once the first good kentriodontid skulls were being discovered there. Part of this mess was mopped up by Barnes and Mitchell in their 1984 reevaluation of Kentriodon obscurus, to which they referred a skull - the species was originally named Grypolithax obscura, which Barnes and Mitchell thought was very close to Kentriodon pernix, and a new skull confirmed the presence of Kentriodon from Sharktooth Hill; the species was recombined and the skull referred to the species. With that overly long intro out of the way, this new paper reports one of the first new kentriodontid specimens from the west coast reported in decades - a well-preserved skull with associated mandible fragments, teeth, vertebrae, and ribs (sadly, no periotic is preserved). The new dolphin, named Kentriodon diusinus, is from the middle Miocene Rosarito Beach Formation near La Mision in northwestern Baja California - a well known locality notable for having a number of species closely related to those from Sharktooth Hill. Kentriodon diusinus is relatively small and has a tapering rostrum with single rooted, conical, polydont teeth, and relatively large and elongate pterygoid sinuses on the basicranium that extend very far anteriorly, lending the dolphin the species name diusinus. Phylogenetic analysis indicates that Kentriodon diusinus is most closely related to the Sharktooth Hill taxon Kentriodon obscurus; interestingly, the specimen has previously been identified in abstracts as Kentriodon sp., cf. K. obscurus. Many more spectacular marine mammal fossils from the Oligocene and Neogene of Baja California await publication, and I'm pleased to see this published.

This study (in Japanese) reports taphonomic details of a whale skeleton collected from the Pisco Formation in Peru. The fossil was originally collected by the Black Hills Institute, and presumably purchased by the Gunma Museum of Natural History; I remember seeing the specimen on the BHI website and hoping it would end up in a museum. The specimen is spectacularly preserved and represents an archaic balaenopterid like Protororqualus or “Balaenopteracortesi var. portisi. Associated with the skeleton are 16 shark teeth all belonging to the extinct “mako” Carcharodon hastalis. Furthermore, the fact that at least two teeth are from the same tooth position, indicating that at least two sharks likely fed on the carcass and shed teeth during feeding.

A partial skull and postcranial skeleton of a small odontocete was found at the Trig Z locality in the Waitaki Valley, a region well known for its upper Oligocene marine vertebrates. This specimen was collected in 1954 by T.G. Marples, the son of Otago Zoology professor Brian Marples (who collected the above mentioned material of Tohoraata waitakiensis; see Boessenecker and Fordyce 2014A, above), and named Prosqualodon marplesi in his honor as part of an honor’s thesis by Mel Dickson in 1964. Historically, Prosqualodon has been a bit of a garbage can; two correctly referred species exist, including Prosqualodon australis from Argentina and Prosqualodon davidis from Australia (the holotype of which has been lost). In the 1930’s, Professor Benham described another odontocete from Otago – probably from the lower Miocene Caversham Sandstone not far off from Dunedin itself (the town where our university is located) – and named it Prosqualodon hamiltoni. This thing is enormous, and does represent a squalodontid, but is clearly much larger than Prosqualodon and had a longer rostrum, and is also needing redescription. Prosqualodon marplesi on the other hand is perhaps more obviously not a squalodontid, and in 1994 R.E. Fordyce recombined it as Notocetus marplesi. Years later, it became evident that this was also incorrect, and as part of Yoshi Tanaka’s Ph.D. thesis he redescribed P. marplesi. Confirming using cladistic analysis that it did not belong in the genus Notocetus, he erected a new genus for it – Otekaikea. Cladistic analysis also supported a sister taxon relationship between Otekaikea and Waipatia, and thus the new genus represents a waipatiid rather than a squalodelphinid (as it was reinterpreted by Fordyce, 1994). Otekaikea and Waipatia are separated only be some rather subtle differences in the braincase and periotic, which I will abstain from recalling to please non-odontocete interested readers. Cladistic analysis also indicates that the Squalodelphinidae – previously assumed to be a clade – is possibly paraphyletic (although see Lambert et al. 2014, who found squalodelphinid monophyly, but used a smaller matrix).

This is yet another report of a new cetotheriid from Paratethys. This new mysticete, Zygiocetus nartorum, was previously referred to as "Cetotherium sp." in an earlier publication by Tarasenko, and is known from similarly aged upper Miocene marine rocks from Adygea (Adygea is a small region of the Russian Federation, and lies to the southeast of Ukraine in the western Caucasus). Most of the features that actually differentiate it from Cetotherium are relatively minor and include subtle differences in the nuchal crests on the braincase and differences in the shape of the periotic. I'm not necessarily certain that these differences merit separation at the genus level. Unfortunately, the figures are somewhat spartan and it's not easy to interpret aspects of the tympanoperiotic morphology. The article oversimplifies patterns of cranial telescoping amongst mysticetes, claiming to identify a new "type" of telescoping; in fact, the different reported patterns have been widely recognized by mysticete workers for nearly a century (to say nothing of the lack of remarks concerned with telescoping in balaenids, eschrichtiids, eomysticetids, and toothed mysticetes).

Ontogeny is a confounding problem in paleontology (see below), but it also can help us understand how certain morphologies evolve. A classic research focus in evolutionary biology is the study of heterochrony – changes in ontogenetic timing. Most evolutionary changes are thought to arise by certain differences in timing. For example, many features evolve simply by appearing slightly earlier during ontogeny during later generations – a classic example would be the gradual loss of abundant body hair in the human lineage. Less hair is common in hominoid juveniles, and in humans the appearance of hair has been delayed later and later into ontogeny, and hairless stages of prenatal ontogeny have been delayed well after birth. Can heterochrony explain some of the morphological diversity and evolution of baleen whales? As part of his Ph.D., my labmate Tsai conducted a geometric morphometric analysis (using relative warps, IIRC) of mysticete ontogeny, focusing on the pygmy right whale Caperea marginata, the sei whale Balaenoptera borealis, and the humpback whale Megaptera novaeangliae. The analysis indicated that only minimal changes occur during the ontogeny of Caperea, with juvenile and adult specimens plotting in the same region of morphospace. However, adult Balaenoptera and Megaptera plotted closely, with juveniles of each plotting close together but not with the adults – indicating not only more extreme ontogenetic changes (generally reflecting the extreme lengthening of the rostrum in rorquals) in rorquals relative to Caperea, but similar ontogenetic trajectories in each rorqual. Tsai and Fordyce thus robustly identify Caperea as undergoing paedomorphosis – retention of juvenile features into adulthood – whereas the two rorquals undergo substantially more changes during ontogeny than the hypothetical ancestral mysticete, and thus undergo peramorphosis.

You and I, and almost all adult humans (barring rare developmental disorders) do not look like we did when we were babies, juveniles, or even awkward teenagers. This is mostly true of all mammals; as mammals mature – some faster than others – a series of skeletal changes accompany all sorts of other soft tissue anatomical and behavioral changes. Many mammals are born with relatively short snouts which then lengthen. In some cases, juveniles of certain species resemble one another closely enough that species differences are not obvious until maturity (a common phenomenon in birds). Biologists have the luxury of tracking a single individual or watching juveniles interact with adults or actually keeping a captive individual to watch it grow and see how it changes; indeed, ontogenetic changes are so easily interpreted that most birding guides will show pictures of juveniles alongside adults for quick visual identification. But paleontologists are not so lucky; most fossils don’t have ancient DNA to sample and confirm relationships, and certainly all fossils are dead, forever. How can we tell a juvenile from an adult? Occasionally size is used – but some paleontologists assume that juveniles will look identical to the adult, and assume a smaller specimen with a few slightly different features from a larger specimen might instead represent a new dwarf species that is closely related. In dinosaur paleontology, a number of paleontologists have gotten into trouble for naming new species that other researchers have claimed actually represent ontogenetic synonyms. How do we assess ontogeny in fossils? We can use paleohistology; others suggest that you can use cladistics, and that if a potential new species plots separately, then it truly is separate – whether juvenile or not. Some work on hadrosaurid dinosaurs has indicated that when known juveniles are coded they still plot out down the tree away from the adults, with other juveniles. In mammals it’s a bit easier because we have teeth, and only one adult set – but baleen whales do not, posing a problem not unlike that facing dinosaur workers. With this in mind, my labmate Tsai took juvenile and adult specimens of modern species of baleen whales and coded them separately and ran it in two different previously published datasets; he used a sei whale (Balaenoptera borealis) and a pygmy right whale (Caperea marginata). Caperea – which is paedomorphic with similar adult and juveniles – plotted together, indicating ontogeny is not much of a confounding factor. However, the juvenile Balaenoptera plotted well outside Balaenopteridae, and outside Crown Mysticeti (Balaenomorpha) in one analysis and with Balaenidae (right whales) in the other. Why? Balaenopterids undergo a number of cranial transformations during ontogeny, and the morphological gulf between juvenile and adult is far, far greater than for the pygmy right whale. Few fossil mysticetes for which fossil ontogenies are known are like Caperea, and thus the potential for being misled by cladistics and misnaming of juveniles as new species in the fossil record is great. Possible remedies include not basing new names on juveniles, unless you can demonstrate that some of the diagnostic features do not change during ontogeny.

Marine mammals from the Pleistocene are rare - for most of the Pleistocene, sea level was lower - meaning that most Pleistocene marine vertebrate bearing localities are below sea level. Occasionally we get areas where dredging yields plenty of marine mammal fossils - and hot spots such as this include the North Sea, offshore Portugal and Spain, offshore South Africa, and the Penghu channel immediately west of Taiwan. Taiwan has a limited but crucial marine mammal record - few other localities are as close to the equator, and it's the lowest latitude marine mammal locality in the western North Pacific; the next richly fossiliferous localities to the south are in Australia and New Zealand. The majority of fossils from the Penghu channel are terrestrial mammals, but various odontocetes have been dredged - and this report describes two immature braincases of gray whales, Eschrichtius sp. Gray whales are perhaps most famous for being commonly visible in the spring along the western US where they are easily seen migrating along the coast (on a drive north from Los Angeles, my wife and I counted 40+ gray whale spouts (aka blow) in under two hours of driving through Big Sur). However, few outside cetology are readily aware that 1) gray whales were declared extinct and rediscovered, 2) gray whales used to exist in the North Atlantic but have been extinct in Europe since about 500 AD and the east coast of North America since about 1700, and 3) still exist in the western Pacific (albeit with a much smaller population). Gray whales undertake marathon migrations from their calving ground in Baja California to the Bering Sea in the eastern Pacific, and from the waters off northern China and Japan to the sea of Okhotsk in the western Pacific. The location of the western population's calving ground is unknown, but thought to be somewhere around the coast of China. The presence of two juvenile specimens of Pleistocene age suggests that perhaps the breeding ground was at least as far south as southern Taiwan, and possibly even further south.

The postcranial skeleton of archaeocetes is famous for demonstrating a beautiful transition from the largely terrestrial mammalian bodyplan to the highly modified skeleton of “modern” cetaceans; key transitions include the loss of weight bearing ability in the fore- and hind-limbs, “locking” of the elbow, modification of carpals into blocky elements and planar interphalangeal and metacarpal joints for stiffening of the carpus, loss of hooflike ungual phalanges, loss of fusion of the sacral vertebrae and decrease in number of functional sacral vertebrae, decoupling of pelvis from vertebral column, reduction of the hindlimb, increase in number of postcervical vertebrae, and modification of the terminal caudal series for the caudal fluke and peduncle. Excellent skeletons of some protocetids exist (Rodhocetus, Artiocetus, Maiacetus) but detailed descriptions have yet to be published; good descriptions exist for the holotype skeleton of Ambulocetus and postcrania of the pakicetids Pakicetus and Ichthyolestes, but these are all isolated elements from a bonebed and some uncertainty exists as to their allocation. Regardless, protocetids represent the crucial stage between mostly terrestrial and fully marine cetaceans. A new skeleton of the protocetid Natchitochia jonesi from the middle Eocene permits a more detailed assessment of the pelvic evolution in protocetids. Described earlier from an associated set of vertebrae, this new specimen is a match but also possesses part of a hindlimb including a well preserved innominate and a femur. The innominate of Natchitochia dwarfs that of Georgiacetus and most other protocetids, and the vertebrae indicate it had two sacral vertebrae (as opposed to four, the primitive condition amongst artiodactyls). In comparison, Georgiacetus is thought to have zero functional sacral vertebrae (like basilosaurids). In 2008 Uhen referred some postcrania to Georgiacetus that suggested that it retained robust hindlimbs but had a pelvis decoupled from the vertebral column; this new specimen of Natchitochia has a robust femur and innominate, but unfused sacral vertebrae with a minimal connection to the pelvis. This new specimen thus supports the idea that protocetids utilized dorsoventral undulations of the hindlimbs rather than the tail for aquatic propulsion – however, see Bajpai and Thewissen above (although, it’s important to note that the protocetid postcrania reported by Bajpai and Thewissen lack hindlimb elements, so it is not objectively possible to compare the relative importance of the tail versus the hindlimb in locomotion – unlike Natchitochia).

Protocetids are a group of archaeocetes that were both arguably the first oceangoing cetaceans, and also the last group capable of hauling out on land – perhaps equivalent to the current stage of evolution of seals and sea lions. Protocetids are well known from the southeastern USA, and include important taxa like Georgiacetus and Carolinacetus, which have informed us greatly on the locomotion and ear anatomy of protocetids. This new paper by Mark Uhen reports several new occurrences of protocetid teeth from New Jersey and South Carolina. Specimens from New Jersey were actually collected from a Miocene unit alongside many obvious examples of Miocene terrestrial mammals, leading Uhen to interpret these as reworked teeth, since protocetids are not known anywhere else to be younger than middle Eocene. These specimens are not anatomically revealing, but importantly the New Jersey specimens reflect the farthest north that protocetids have been discovered. Uhen suggested that protocetids were confined to the Tethys and north Atlantic owing to their absence in other well-sampled Eocene rocks from other regions. However, further sampling in places like Peru, New Zealand, and Seymour Island (Antarctica) is probably necessary. Several protocetids were reported from fragmentary remains by Uhen et al. in 2011 but are apparently now regarded as basilosaurids (they are similar to “Eocetus”, one species of which – Eocetus wardii – was transferred to the new archaic basilosaurid genus Basilotritus last year).

This solid new study by Jorge Vélez-Juarbe and Daryl Domning reports a new Oligocene sirenian from the southeastern United States, in the ninth paper in the long-standing Atlantic and Caribbean fossil sirenian series. The genus Metaxytherium is a speciose grade of sea cows formerly known only from the Miocene and earliest Pliocene. Fossils of Metaxytherium are known from all over the world, including the western and eastern North Atlantic, the eastern North Pacific, the Mediterranean, Indian Ocean, and western South Atlantic. The Metaxytherium “lineage” is paraphyletic, and the Hydrodamalinae (Dusisiren + Hydrodamalis – the giant sea cows) is nested within (and of course, the genus Dusisiren is just as paraphyletic). The new species, Metaxytherium albifontanum, extends the fossil record of the genus back into the Oligocene. This new species is rather small, only 2-3 meters in length, and is known from three well-preserved partial skeletons from the Parachucla Formation of Florida and the Chandler Bridge Formation of South Carolina (the latter is the same unit which produced the holotype specimens of notable fossil cetaceans such as Agorophius, Cotylocara, Eomysticetus, Micromysticetus, and Xenorophus). Interestingly, this new species of Metaxytherium was apparently sympatric with other sirenians such as Crenatosiren olseni and Dioplotherium manigaulti, and because it had smaller tusks and a less deflected rostrum, M. albifontanum likely relied less upon uprooting rhizomes from the sediment than these other sirenians. Lastly, this new species suggests that a west Atlantic origin for the Metaxytherium lineage is likely, rather than a Tethyan origin.

And this study marks the tenth contribution in the long standing Atlantic and Caribbean Sirenia series, and Jorge’s second installment as first author. This study reports a larger sirenian, the holotype of which was collected by none other than the first author himself in 2006 and 2007 while doing fieldwork in Puerto Rico. I remember seeing occasional photos of the specimen as Jorge was preparing it; it is more than admirable for a paleontologist to have the necessary skills to excavate, prepare, and publish their own fossils (in fact, that is the very definition of a paleontologist, and unfortunately many researchers do only the latter). The type specimen of Priscosiren atlantica is from the San Sebastián Formation and is early Oligocene in age; several additional specimens were referred from the same unit, and critically, also from the Ashley Formation of South Carolina. Priscosiren is a halitheriine dugongid, and other cladistic analyses by Jorge and Daryl have confirmed its position as a basal dugongid. Priscosiren was likely sympatric with other Oligocene sirenians such as Caribosiren and Crenatosiren (being somewhat older than the late Oligocene sirenian assemblage from the Chandler Bridge Fm. – e.g. Metaxytherium, Crenatosiren, and Dioplotherium). Priscosiren had a poorly deflected rostrum with small tusks (not dissimilar from Metaxytherium albifontanum), and as in all known sirenian assemblages there are subtle differences in feeding morphology between each sirenian present – but without a clear, consistent pattern through earth history, suggesting these are “dispersal assemblages” or represent “chance structuring”. Another solid piece on sirenians, with beautiful color images to boot!

Morocco is rather famous for a series of spectacular marine vertebrate fossil assemblages spanning the late Cretaceous through Eocene – many fossils of mosasaurs, and Paleocene-Eocene shark teeth are easily found for sale. In fact, K-Pg shark teeth from Morocco can be found in virtually every rock shop in the United States, and in every country I’ve ever been in, including the UK and New Zealand, I have seen them for sale. Normally when abundant sharks are found, other marine vertebrates are easily recovered; while the Moroccan record of mosasaurs is great for the Cretaceous, but unfortunately not much in the way of marine mammals have been found in Eocene rocks. My colleague Rachel Racicot (Postdoctoral fellow at Howard U.) went to these coastal localities with archaeocete specialist Mark Uhen (George Mason U.) and didn’t turn up much; the localities are pretty heavily picked over and quarried by commercial collectors. Still, some tantalizing bits have been found, and this new paper reports some of the first Eocene marine mammal assemblages from Morocco. Cetaceans include a host of basilosaurid archaeocetes, including cf. Saghacetus sp. (rostrum fragment with teeth, vertebrae), cf. Stromerius sp. (vertebrae, humerus), cf. Dorudon atrox (vertebrae), cf. Dorudon sp. (vertebrae, phalanx), and Basilosaurus isis (partial articulated vertebral column); the sirenian cf. Eosiren sp. (ribs) is also present. Fossil vertebrate material is concentrated into two bonebeds separated by a few meters of rhythmic estuarine deposits. The two bonebeds are for some odd reason correlated with two sea level low stands – which is curious, since marine bonebeds are typically formed during transgressions (e.g. see Boessenecker et al., above), and both bonebeds mark deposits deeper than the intervening estuarine deposits.