All toothless whales are baleen whales

Updated November 20, 2016 based on input from Dr. RW Boessenecker who corrected a mistake I made reconstructing Tokarahia by noting the mandible was axially rotated in situ 180º. Correcting that error produces the bowed out Gothic arch set of mandibles, typical of mysticetes. 

Preamble
Several of the mistakes discussed below are based on the long-standing tradition that whales are monophyletic. That is, odontocetes (toothed whales) and mysticetes (baleen whales) were long thought to have a common ancestor with flukes and fins. That paradigm  was overturned by the large reptile tree (LRT) earlier here. Workers thought that common ancestor would be an archaeocete, but, so far, all tested archaeocetes nest basal to extant odontocetes and were derived from tenrecs in the LRT.

Almost fifty years ago whale monophyly was questioned
by Van Valen (1968), who listed a number of traits that distinguish Odontoceti from Mysticeti. Unfortunately this was before computer-assisted phylogenetic analysis and neither desmostylians nor tenrecs were offered as basal taxa with legs. This was also long before any whales with legs had been discovered. Gotta give Van Valen credit for his insight way back then.

Yesterday we looked at the desmostylian ancestors of today’s mysticete (baleen) whales. Less than 24 hours ago I encountered for the first time Aetiocetus (Emlong 1966; Figs. 1, 2). I learned it has long been considered a toothed basal mysticete. Evidently some of the back teeth are leaf-shaped and all of the teeth are small and widely spaced. Most whale workers are happy with this hypothesis or relationships, but the LRT finds otherwise based on an expanded taxon list.

Figure 1. Aetiocetus skull in several views.

Figure 1. Aetiocetus skull in several views. Most whale workers today consider this taxon close to the origin  of baleen whales. The transversely crested cranium is a trait found in living odontocetes, not mysticetes.

I added Aetiocetus
to the large reptile tree and, while given the opportunity to nest with mysticetes, Aetiocetus nested instead between Zygorhiza and Orcinus + Physeter, all members of the Odontoceti. As recently as 2015, Ekdale et al. (Fig. 2) were trying to use Aetiocetus to explain the origin of baleen in modern whales.

Figure 2. Palate and teeth of the odontocete Aetiocetus alongside palates of juvenile gray whale and embryo fin whale, members of the Mysticeti. Aetiocetus was long thought to be a basal mysticete.

Figure 2. Palate and teeth of the odontocete Aetiocetus alongside palates of juvenile gray whale and embryo fin whale, members of the Mysticeti. Aetiocetus was long thought to be a basal mysticete.

18 hours ago I encountered another mysticete,
Tokarahia kauaeroa (Boessenecker and Fordyce 2015; Late Oligocene; OU 2235), which has no teeth. Their reconstruction (Fig. 2) is

Figure 2. Tokarahia, a toothless odontocete long thought to be a basal mysticete. Original interpretation of materials is presented alongside a new interpretation, closer to the bones in situ. See figure 4.

Figure 3. Tokarahia, a toothless odontocete long thought to be a basal mysticete. Original interpretation of materials is presented alongside a new interpretation, closer to the bones in situ. See figure 4. The humerus is rotated so the ball joint fits into the ventral socket of the scapula.

Unfortunately
I thought Boessenecker and Fordyce changed the curve of the mandible in their reconstruction and did not follow the very narrow mandibles in restoring the largely missing or buried rostrum. They also moved the orbit anteriorly. I attempted a mistaken correction (Fig. 3) that was based on a narrow mandible interpretation. Dr Boessenecker reported the mandible was axially rotated in situ (Fig. 4). Those corrections was applied shortly thereafter. A good lesson in keeping an open mind.

Figure 1. Tokarahia in situ and as originally reconstructed (on right). Flipping the right mandible and reconstructing the skull anew (at left).

Figure 4. Tokarahia in situ and as originally reconstructed (on right). Flipping the right mandible and reconstructing the skull anew (at left).

12 hours ago
I also learned about Isanacetus laticephalus (Kimura and Ozawa 2002; early Miocene, 18 mya; MFM 28501; Fig. 5, 6). This fossil whale is indeed a mysticete. In the LRT it nests between the desmostylian Behemotops (presumably with at least front legs) and extant baleen whales.

Figure 5. Isanacetus skull in several views. I also present skull tracings in DGS that differ in some respects from the published drawings.

Figure 5. Isanacetus skull in several views. I also present skull tracings in DGS that differ in some respects from the published drawings. Isanacetus is a basal mysticete, derived from a sister to Behemotops.

Like all mysticetes and derived desmostylians
Isanacetus has a ventrally concave rostrum, a wide flat skull and other traits that distinguish mysticetes from odontocetes + tenrecs. Isanacetus is also one of the smallest known mysticetes, about twice the size of its current desmostylian sister, Behemotops (Fig. 6) and half again longer in the skull than the more completely known Desmostylus.

Figure 6. Isanacetus compared to sisters recovered in the LRT. Balaeonoptera is much reduced.

Figure 6. Isanacetus compared to sisters recovered in the LRT. Balaeonoptera is much reduced. The loss of teeth actually occurred when mysticetes had legs!

More backstory for those keenly interested
Below are some earlier and traditional reports on Aetiocetus and aetiocetids, which was considered by these authors to be related to mysticetes and cetotheres (extinct mysticetes).

Whitmore and Sanders 1976 report,
“The cheek teeth [of Aetiocetus] are leaf shaped, similar to those of Patriocetus, but smaller and with the roots coalesced. The triangular rostrum, reduced dentition, and the conformation of the posterior ends of the maxillae, premaxillae, and nasals (Ernlong, 1966:s) are characters that would be expected in the ancestor of the mysticetes. Thenius (1969:489) stated: “Even if Aetiocetus, because of its geologic age (upper Oligocene) cannot be a direct stem form of the cetotheres, yet this genus documents that a specific family (Aetiocetidae) must be classified as ancestor, the link between ancient and baleen whales.

“Among the few Cetacea known from deposits of middle Oligocene age are two occurrences of unmistakable Mysticeti. One of these, Mauicetus Benham, 1939, from New Zealafid, has long nasals embraced by premaxillae and maxillae which extend posteriorly to the level of the supraorbital process of the frontal, together with an anteriorly thrusting triangular supraoccipital. The Oligocene Mysticeti, had already evolved the elongated, edentulous rostrum, constituting 3/4 to 4/5 of total skull length, that typifies the modern baleen whales. The mandible of Oligocene Mysticeti was also edentulous and, like those of modem baleen whales, was long and slim.”

Berta and Demere 2005 reported,
“Aetiocetids are the most taxonomically and morphologically diverse clade of toothed mysticetes known from the late Oligocene of the eastern and western North Pacific. Aetiocetids can be distinguished from other toothed mysticetes by the following unequivocal synapomorphies: lobate or triangular parietal-frontal suture; zygomatic process of squamosal expanded near anterior end; “window” in the palate exposing vomer; short, broad extension of the palatine that overlaps the pterygoid; and exoccipital developed ventrally as an anteriorly directed posterior sinus.

The presence of palatal nutrient foramina associated with the upper teeth in all aetiocetids suggests that these toothed mysticetes had already evolved some type of baleen. The form and function of this rudimentary baleen is currently unknown, but the fact that these archaic mysticetes also possessed procumbent anterior teeth, broad diastemata, and posterior teeth with sharply pointed cusps, accessory denticles, and longitudinal enamel ridges suggests development of a specialized type of filter feeding differing from that of other toothed and edentulous mysticetes.”

Unfortunately
these authors had not expanded their taxon set to include desmostylians, which pull mysticetes away from odontocetes + tenrecs.

Eckland et al. 2015 report
“The origin of baleen in mysticetes heralded a major transition during cetacean evolution. Extant mysticetes are edentulous in adulthood, but rudimentary teeth develop in utero within open maxillary and mandibular alveolar grooves. The teeth are resorbed prenatally and the alveolar grooves close as baleen germ develops.”

That’s all well and good, but you really need a wide gamut taxon inclusion set that includes whales, tenrecs and desmostylians to find the diphyletic origins of extant whales and in desmostylians one should look for the origin of baleen, as discussed earlier here.

References
Berta A and Demere TA 2005. Phylogenetic relationships among the diverse toothed mysticete clade the aetiocetidae and reconsideration of the filter feeding niche. Evolution of  aquatic tetrapods. Fourth triennial convention abstracts May 16-20 2005, Akron, OH, USA.
Boessenecker RW and Fordyce RE 2015. A new genus and species of eomysticetid (Cetacea: Mysticeti) and a reinterpretation of ‘Mauicetus’ lophocephalus Marples, 1956: Transitional baleen whales from the upper Oligocene of New Zealand. Zoological Journal of the Linnean Society. in press. doi:10.1111/zoj.12297.
Demere TA 2005. Palate vascularization in an Oligocene toothed mysticete (Cetacea: Mysticeti): Aetiocetidae); implications for the evolution of baleen. Evolution of  aquatic tetrapods. Fourth triennial convention abstracts May 16-20 2005, Akron, OH, USA.
Ekdale EG, Demere TA and Berta A 2015. Vacularization of the gray whale palate (Cetacea, Mysticeti, Eschrichtius robustus): soft tissue evidence for an alveolar source of blood to baleen. The Anatomical Record Advances in Integrative Anatomy and Evolutionary Biology 298(4) · February 2015.
Whitmore FC Jr and Sanders AE 1976. Review of the Oligocene Cetacea. US Geological Survey Staff — Published Research. Paper 237.
Van Valen L 1968. Monophyly or diphyly in the origin of whales”. Evolution. 22 (1):37–41.

wiki/Aetiocetus
wiki/Tokarahia
wiki/Isanacetus – not created yet

20 thoughts on “All toothless whales are baleen whales

  1. You might want to actually fly down to New Zealand and look at the Tokarahia holotype for yourself. And, while you’re at it, I implore you to go look at some actual modern baleen whale skeletons as it might help you better (or, even properly) interpret the weird world of whale anatomy. The right mandible of Tokarahia kauaeroa (OU 22235) is flipped 180 degrees about its long axis so that you’re looking at the ventral edge – the mandible is bowed laterally, just like all Chaeomysticeti… not medially bowed. Modern mysticetes with bowed mandibles (except, perhaps balaenids) have their jaws rotated dorsomedially in lateral view when the mouth is closed, even if in true lateral view the dorsal margin is slightly concave (as in Tokarahia).

    It’s also pretty clear from your “revised reconstruction” (esp. lateral view) that you have no idea what other eomysticetids look like. The skull of Tokarahia is far from the most complete eomysticetid (owing to Osedax bioerosion) and I based missing parts off of Yamatocetus canaliculatus from Japan, the holotype of which has a virtually complete skull. You literally do not have wiggle room to make this thing into…whatever your reconstruction is supposed to be.

    • Thank you for your corrections, Dr. Boessenecker. I have corrected this blog post and its imagery. And to your point, today’s blog post (11-20-2016) includes additional related mysticetes. Every taxon I look at is new to me. That can be a problem, but a correctable one. Good Science admits errors and fixes them.

  2. Is there a reason you disregarded the massive amount of molecular evidence conclusively showing whales to be monophyletic? Just curious.

    • I only wish I could report that the topology of the morphological tree was identical to DNA trees or traditional trees. Because they don’t match, I feel it is good Science to report my findings, especially since all tested sister taxa actually look like they should be sister taxa. That does not always happen with DNA tests or certain traditional analyses that improperly exclude or include certain taxa. If you find any mistakes, or know of any taxa that will shift the topologies back to those you consider valid, please send them to me. I often find errors in the work and make all the corrections I am aware of.

      • I would argue that, given the mountain of molecular evidence demonstrating (conclusively) that Cetacea is monophyletic and that none of its subordinate clades is related to tenrecs, the results you’re getting are suspect at best. A good scientist would have checked that out before reporting his results.

  3. Since this is Science, all results should be repiicable. Unfortunately I don’t have the ability to do DNA testing. I encourage you to put Tenrec next to Pakicetus on the same page or same monitor. I think you’ll find they were ‘separated at birth.’ Or… Perhaps try a morpho test yourself with the taxa I’ve tested using your own characters before you condemn the results. After all, DNA testing is lumping turtles deep into non-turtle territory. We’ve also lived a little too long with golden moles lumping with elephants and other odd ‘sisters’. In the LRT both are separated by a long list of non-Afrothere taxa. In the LRT you find a gradual accumulation of derived traits, which is exactly what we’re looking for when modeling evolutionary events. DNA does not always gives us that across taxa. But it is great when it comes to criminals and families that cross oceans.

    • I stand by what I said. Given the amount of data currently available in the literature, your conclusion that whales are not monophyletic is wrong.

      • I also disagree with your assertion that you cannot “do DNA testing.” You may not be able to do the sequencing yourself, but the sequence data (both aligned and unaligned) are usually available, from the literature if not from the authors. Software applications to align and phylogenetically analyze the data are easily available, as are books that will guide you through the basic steps of doing it.

    • Searches in scholar.google.com should bring up several. But that aside, and given what you wrote about molecular phylogenetics just now, I think you should consult some of the more general references on the subject first.

  4. I didn’t write anything about molecular phylogenetics. I wrote about resulting tree topologies that don’t make sense when they come out of molecular data. Frustrating, isn’t it… when cherished institutions and paradigms fall short. It’s okay. It will take awhile to sink in. Part of the deal.

  5. I only wish critics of the results would 1) be more specific, less vague; 2) test the results themselves before denying their validity; 3) look at the taxa involved and wonder why such similarities were previously overlooked; and basically, act like a scientist. Others say, I need more characters, but how many and which ones will change the tree topology? You say I need to study molecular phylogenetics, but how much and which resources will change the tree topology. Chris, if you see any taxa that do not resemble their sisters and you know where they should nest instead, please let me know. We’ll start there. Right now this tree is fully resolved as is and it tells a great story that makes sense. Other trees do not make sense or as much sense across their entire breadth.

    Flat denial is form of blackwashing that is beneath anyone with a PhD. When I referee a paper, I include specific margin comments. I expect my referees to do the same.

    And you should know that several recent ‘discoveries’ by others were made several years ago here. So there is a growing record of confirmation building on the present study.

    • So you’re saying the methodology doesn’t matter if you don’t like the results?

      I’m not a mammal expert, but I know about different kinds of data, the strengths and weaknesses that come with each, and the different kinds of methods that can be applied to them. Taxon sampling is an issue for everyone, but it doesn’t affect molecular data in the same way it affects morphological data. And the most recent molecular analyses have sufficient taxon sampling to demonstrate (a) afrotherian monophyly and (b) cetacean monophyly.

      Rather than simply dismissing it because you dislike the results, why don’t you accept my invitation to learn more about molecular phylogenetics?

  6. It’s not that I dislike DNA results. They don’t include fossils. And they don’t match morphological studies that do include fossils. And the results should match, if valid. Everything should stand up to testing. DNA results over large phylogenetic distances, at present, do not. I’m not sure what’s wrong with them, but I will accept your invitation to study what is available online over this long weekend. Like Galileo, I trust experiments and results I can see and do. And others can repeat. I hope that earns your respect.

    • “Online over the long weekend” won’t really be enough. We teach whole courses on the subject. I would strongly recommend one of the excellent textbooks that you can purchase. Limey and Salemi’s Phylogenetic Handbook would be a great one to start with.

      I’ve seen several places where you illustrate a serious misunderstanding of how molecular analyses are done, and what they’re supposed to accomplish. For example, you state that molecular analyses don’t use plesiomorphic outgroups. But “plesiomorphic” doesn’t necessarily mean anything in molecular biology; two sister groups will have been separate for the same length of time, and both will have been evolving throughout that time. Hence, outgroups are chosen more for accurate rooting than for the replication of an ancestral condition.

      You’ve also claimed that molecular trees are problematic because recovered sister taxa don’t always look alike. Well….no, they don’t. Nor should they, if they’ve been separate lineages for long enough. Afrotheria looks odd when you only consider a golden mole and an elephant, but these have been separate lineages for anywhere between 60 and 90 million years – so no one expects them to look similar.

      Bear in mind – I’m a morphologist. I do work with molecular data, but I mostly work with morphology. And increasingly, molecular and morphological analyses do, indeed, give very similar results. They match more often than they don’t. Those that appeared to clash in the past usually involved improper comparisons (e.g. a molecular analysis vs. the taxonomy in a field guide that might have been done in the 1950’s), a rooting problem, or a handful of branches amid an overall picture of congruence. That’s been the case with the group I know best (crocodylians); gharials are an ongoing problem, but if you take one of them out, you basically get the same tree regardless of which data set you’re using.

      Please – read up on the subject (more than just “online”).

  7. I appreciate your ‘faith’ in the unseen (the last common ancestor of golden moles and elephants), but after testing them, I find more parsimonious nesting partners that greatly separate putative relatives recovered by molecules. And the ‘ghost’ lineages are much shorter. The similarities of these new sisters goes from head to toe. And that goes for pterosaurs, mesosaurs, turtles and the entire list of former enigma taxa. You might ask yourself, why does my tree nest the taxa together that everyone agrees on, but finds new nesting partners for the controversial ones? If the methodology was wrong, there should be some odd nestings in the list that you could easily call out.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s