Stem taxa: wider vs. narrower definitions

The following is as much a learning experience for me
as it may be for you, as I come to an understanding of a definition with which I was previously unfamiliar — Stem Taxa in the ‘wider sense’. There is also a ‘narrower sense’, with which I was more familiar (see below).

Stem taxa in the wider sense
are extinct taxa closer to one clade of living taxa than to any other living taxa. In more precise terms, according to Wikipedia: “A stem group is a paraphyletic group composed of a pan-group or total-group, above, minus the crown group itself (and therefore minus all living members of the pan-group).”

Figure 1. CLICK TO ENLARGE. Stem taxa are closest ancestors to living taxa. Here basal diapsids and marine enaliosaurs are stem archosaurs. Triceratops is a stem bird. Captorhinids are stem turtles. Pterosaurs are stem squamates.

Figure 1. CLICK TO ENLARGE and retrieve a PDF file. Stem taxa are the closest ancestors to living taxa, but not the living taxa. Here in the wider sense marine enaliosaurs, including ichthyosaurs, are stem archosaurs. Triceratops is a stem bird. Diadectids are stem turtles. Pterosaurs are stem squamates.

The stem reptiles
in the LRT (Fig. 1, click here for PDF enlargement) are not the same stem taxa found in traditional studies, like Benton 1999, who includes pterosaurs among the stem birds. In the LRT (Fig. 1) pterosaurs are stem squamates, not squamates, but closer to squamates than to Sphenodon, the extant tuatara. Stem amphibians are not listed here, because the only listed extant taxon in that clade is Rana, the bull frog. Here plesiosaurs, ichthyosaurs and mesosaurs are all stem archosaurs. All dinosaurs are stem birds. Diadectids are stem turtles.

Paleontological significance
According to Wikipedia, “Placing fossils in their right order in a stem group allows the order of these acquisitions to be established, and thus the ecological and functional setting of the evolution of the major features of the group in question. Stem groups thus offer a route to integrate unique palaeontological data into questions of the evolution of living organisms. Furthermore, they show that fossils that were considered to lie in their own separate group because they did not show all the diagnostic features of a living clade, can nevertheless be related to it by lying in its stem group.” Unfortunately, these benefits get fuzzier as the phylogenetic distance increases.

Stem group: the narrower sense
According to Wikipedia (referencing Czaplewski, Vaughan and Ryan 2000), “Alternatively, the term “stem group” is sometimes used in a narrower sense to cover just the members of the traditional taxon falling outside the crown group. Permian synapsids like Dimetrodon and Anteosaurus are stem mammals in the wider sense but not in the narrower one. From a cynodont ancestry, the stem mammals arose in the late Triassic, slightly after the first appearance of dinosaurs.”

That narrower definition
is the one I was following and is certainly more useful, more targeted, etc. Since both definitions are in play in the wider world, be sure you specify which one you are discussing. Dr. Naish embraced the wider one while ignoring the narrower one.

And that’s okay.

References
Benton MJ 1999. Scleromochlus taylori and the origin of the pterosaurs. Philosophical Transactions of the Royal Society London, Series B 354 1423-1446.
Czaplewski TA, Vaughan JM and Ryan NJ 2000. Mammalogy (4th ed.). Fort Worth: Brooks/Cole Thomson Learning. p. 61.

wiki/Crown_group#Stem_groups

 

Advertisements

Whales are diphyletic. Goodbye ‘Cetacea’!

Updated November 8. 2016 with a revision to the nesting of Janjucetus after the inclusion of the desmostylian and mysticete sister, Behemotops.

You might have suspected this all along
since the two orders of whales, Odontoceti and Mysticeti, are so different from one another. The present study indicates we will never find a last common ancestor with flippers and fins for all living whales because the two orders arose from two widely separated mammal orders with legs.

Baleen whales (Mysticeti) arise from Paleoparadoxia and Anthracobune in the large reptile tree. So there IS now evidence in the LRT for a hippo relationship, as these taxa are more or less related to hippos. But the hippo clade (Mesonychidae) is separated from the artiodactyls by several nodes, so even mysticete whales are not close to artiodactyls.

Toothed whales (Odontoceti) still arise from tenrecs as noted here earlier.

That makes living whales diphyletic
and the order name ‘Cetacea’ invalid and useless because it is not monophyletic. It will probably take a century to disappear because it is so ingrained.

DNA tests validated
at least for Mysticeti as they relate to hippos, the only living mesonychids. Not sure if Odontoceti has been tested against tenrecs in DNA testing.

Discovered during testing
When you include living odontocetes and mysticetes in the same cladogram these taxa will nest together with either tenrecs or mesonychids in two separate trees due to their massive convergence. However, when you enter them in separate tests (Fig. 1) the two orders of whales will nest separately based on the present taxon and character lists.

Figure 5. Subset of the LRT, higher placental mammals with a focus on whales (yellow) and their ancestral clades, the Tenrecidae and Mesonychidae. Both are a fair distance from artiodactyls.

Figure 1. Subset of the LRT, higher placental mammals with a focus on whales (yellow) and their ancestral clades, the Tenrecidae and Mesonychidae. Both are a fair distance from artiodactyls. This cladogram updates an earlier one that nested Janjucetus with Balaenoptera.

This we learn
after adding Janjucetus hunderi (Fig. 2; Fitzgerald 2006; Late Oligocene; NMV P216929 (Museum Victoria Palaeontology Collection, Melbourne, Australia), Orcinus (killer whale; Fig. 5) and Balaenoptera musculus (blue whale; Fig. 4) to the large reptile tree (Fig. 1).

Figure 1. Janjucetus is known from a well-preserved skull, a scapula rib and not much else. Here it is compared to Paleoparadoxia to scale. Janjucetus mya have had legs, or leg/flippers.

Figure 2. Janjucetus is known from a well-preserved skull, a scapula rib and not much else. Here it is as a whale compared to Paleoparadoxia to scale. Janjucetus had legs, not flippers and flukes based on phylogenetic bracketing.

That low wide skull
and procumbent premaxillary teeth of Janucetus (Fig. 3) are obvious early clues that we’re not dealing with giant carnivorous tenrecs here, but with herbivores at home underwater, like hippos and desmostylians (including Paleoparadoxus (Fig. 2). Baleen whales also lack echolocation skills, something toothed whales got from tiny tenrecs.

Wikipedia reports
“Desmostylians are the only known extinct order of marine mammals.”
Not any more.

From the Fitzgerald abstract
“Unlike all other mysticetes, this new whale was small, had enormous eyes and lacked derived adaptations for bulk filter-feeding. Several morphological features suggest that this mysticete was a macrophagous predator, being convergent on some Mesozoic marine reptiles and the extant leopard seal (Hydrurga leptonyx).”

That’s because Janjucetus is not a whale, but a whale ancestor more closely related to Anthracobune (Fig. 1).

Figure 2. The skull of Janjucetus in 3 views. Images from Fitzgerald 2006. Colors added.

Figure 3. The skull of Janjucetus in 3 views. Images from Fitzgerald 2006. Colors added. This is an incredible find that I just yesterday became aware of.

From the Fitzgerald text
“The origins and early evolution of the two extant suborders of Cetacea (Odontoceti and Mysticeti) remain poorly understood. Among the latter two groups, the origins of the Mysticeti (baleen whales), and their unique feeding mechanism,have proved particularly baffling.In this study, I report a new Late Oligocene toothed mysticete from Australia that is more basal than any previously described.”

Hate to report this, but it’s anthracobune, not a whale. The desmostylian, Behemotops makes a better  (more parsimonious) ancestor to baleen whales.

Figure 4. Blue whale (Balaenoptera musculus) skull and skeleton. Note the lack of a thumb goes back to Mesonyx and Paleoparadoxia

Figure 4. Blue whale (Balaenoptera musculus) skull and skeleton. Note the lack of a thumb goes back to Mesonyx and Paleoparadoxia

Unfortunately
Fitzgerald did not test Janjucetus and other whales against any of the taxa in the tenrec clade. Hippopotamus was the outgroup taxon. Has the monophyly of the Cetacea ever been questioned or tested before?

Figure 4. The killer whale (Orcinus orca) skeleton and skull with parts colorized.

Figure 5. The killer whale (Orcinus orca) skeleton and skull with parts colorized. Note the retention of five fingers and a sternal series. This, too, is a giant aquatic tenrec.

A few days ago
we looked at the double-pulley shape of the artiodactyl and whale astragalus (ankle bone). Today I added an image of the Leptictidium (tenrec) ankle, also sharing a double-pulley ankle joint. You can see it here.

Figure 7. Sperm whale (Physeter macrocephalus) with bones colorized for clarity. Note the five fingers and fewer dorsal ribs.

Figure 7. Sperm whale (Physeter macrocephalus) with bones colorized for clarity. Note the five fingers and fewer dorsal ribs.

Whales and molars
In the very little time that I’ve looked at whales and their ancestors it appears that the molars shrink and disappear, losing cusps and size, making room for more canines and more canine-like premolars. The incisors are missing from the premaxilla of the killer whale. Teeth are morphologically plastic and change their shape, size and number often, not only in mammals, but in many other reptile clades. I’m not a whale dentist, so if one is out there, please send any data you may have on whale tooth identification and eruption patterns and I will share it.

And finally,
as you probably know, dugongs are whale-like herbivores with fins and whale-like flukes. So the separation of odontocetes from mysticetes should come as less of a shock. If I’ve made any errors here, please let me know. I’m still a newbie at whale morphology.

References
Fitzgerald EMG 2006. A bizarre new toothed mysticete (Cetacea) from Australia and the early evolution of baleen whales. Proceedings of the Royal Society B 273:2955-2963.

Docodon and molar count

The genus Docodon (Marsh 1881; Late Jurassic, 170 mya; Fig. 1) is represented by a jaw with several more molars than typical (Fig. 1). Hard to tell the premolars from the molars in lateral view. See the dorsal view for the distinct difference. Even so, the count may be off, because molars are not molars based on shape, but on the fact that they appear once and are not replaced during growth. I cannot tell, note have I found references that say where the division is in Docodon.

Figure 1. The holotype of Docodon has 4 incisors, 4 premolars and a whopping 7 molars.

Figure 1. The holotype of Docodon has 4 incisors, 4 premolars and a whopping 7 molars. diplocynodon has 8 according to Osborn, who confirms the 7 in Docodon.

More than four molars in a mammal jaw is relatively rare.
But not rare in ancestral monotremes. Among tested taxa  Amphitherium, Kuehneotherium and Akidolestes have 6.

Figure 1. The addition of teeth in Kuehneosaurus and Akidolestes led to the loss of teeth in Ornithorhynchus.

Figure 1. The addition of teeth in Kuehneosaurus and Akidolestes led to the loss of teeth in Ornithorhynchus.

There is also “a rule” that says
only one canine appears, but the other teeth can vary greatly in number. I’m wondering if that is true. Sometimes there will be a small, simple tooth arising between the canine and the double-rooted premolars. Is that a tiny canine? or a tiny premolar? Maybe someone out there has not only the answer, but the reason why.

And yes,
I’m aware of the convention that numbers premolars 1-4. But the anterior one, is almost always the smallest, as if it just arrived.

References
Marsh OC 1881. Notice of new Jurassic mammals. American Journal of Science. (3) xxi: 511-513.

wiki/Docodon

 

The Eutriconodonta is also paraphyletic in the LRT

Triconodon, Docodon, and Kuehneotherium are known form dentary bones with most of their teeth in place. Generally I avoid adding such partial specimens to the large reptile tree (LRT-updated at 848 taxa) because so few scores are generated for them with the current character list that they lead to loss of resolution at their nodes…

But curiosity won out
when wondering about members of the putative clade Eutriconodonta (Kermack et al. 1973), a clade that ostensibly replaces the paraphyletic Triconodonta. Some of these mandible-only taxa I added to the tree only to delete them later, just to see where they nested (often at the base of the Monotremata, as one would guess given their Mid-to Late Jurassic ages).

According to Wikipedia
“The Eutriconodonta  is a [presumeably monophyletic] order of mammals” broadly, though not exclusively characterized by molar teeth with three main cusps on a crown that were arranged in a row.  “Eutriconodonts retained classical mammalian synapomorphies like epipubic bonesvenomous spurs and sprawling limbs. Eutriconodonts had a modern ear anatomy, the main difference from therians being that the ear ossicles were still somewhat connected to the jaw via the Meckel’s cartilage.

“Phylogenetic studies conducted by Zheng et al. (2013), Zhou et al. (2013) and Yuan et al. (2013) recovered monophyletic Eutriconodonta containing triconodontids, gobiconodontids, Amphilestes, Jeholodens and Yanoconodon. The exact phylogenetic placement of eutriconodonts within Mammaliaformes is uncertain.”

“Traditionally seen as the classical Mesozoic small mammalian insectivores, discoveries over the years have ironically shown them to be among the best examples of the diversity of mammals in this time period, including a vast variety of bauplans, ecological niches and locomotion methods.”

Traditional Eutriconodont taxa (see Martin et al. 2015) presently included in the LRT nest in a variety of clades:

  1. Gobiconodon – pre-mammal tritylodontid
  2. Repenomamus – pre-mammal tritylodontid
  3. Spinolestespre-mammal tritylodontid
  4. Jeholodens – pre-mammal tritylodontid
  5. Volaticotherium basal placental 
  6. Triconodonbasal monotreme
  7. Trioracodon – basal monotreme
  8. Yanoconodonpre-mammal (Fig. 1).

We’ve seen this sort of splitting
of traditionally established clades based chiefly on tooth traits before with the Docodonta (Fig. 1). As in molecule trees, tooth trees are not replicated in the LRT, which recovers a distinctly new tree topology without the odd logic jumps that traditional clades, like Afrotheria, produce.

Figure1. Repeat of an early subset of the LRT, this time highlighting putative eutriconodonts and where they nest. No wonder they are described as a diverse clade!

Figure1. Repeat of an early subset of the LRT, this time highlighting putative eutriconodonts and where they nest. No wonder they are described as a diverse clade!

For the most part
eutriconodonts nest more or less together very close to the base of the Mammalia, whether in or out. Those slender posterior jaw bones are not typically preserved, but the long groove in which they are attached is typically preserved. Caution must be exercised, as fully mammalian taxa like Monodelphis, can also preserve a remnant of this groove despite the complete evolution of the post-dentary bones into tiny ear bones.

Figure 4. Mondelphis domestics with its posteromedial jaw groove highlighted in red. The ear bones are tiny and enclosed within the auditory bulla beneath the cranium.

Figure 2. Mondelphis domestics with its posteromedial jaw groove highlighted in red. The ear bones are tiny and enclosed within the auditory bulla beneath the cranium.

So.. about those venomous ankle spurs…
Ornithorhynchus, the platypus, has them and likely so do its sisters. The authors of the Volaticotherium paper make no mention of either venom nor spur and score it as a “?”. The Yanoconodon tarsi are a challenge to reconstruct based on their preservation. The authors and yours truly note no spur-like bones present.

A new evolution website has launched
Check out www.TimeTree.org for a tremendous amount of phylogenetic information. For instance, one can input two well-known taxa, like Gallus and Homo, and the tree will determine the estimated date of their last common ancestor, in this case Vaughnictis (which is a taxon not in their current database).

References
Editors: Carrano MT et al. 2006. Amniote Paleobiology: Perspectives on the Evolution of Mammals, Birds and Reptiles. University of Chicago Press.  online here.
Kermack KA, Mussett F, Rigney HW 1973. The lower jaw of Morganucodon. Zoological Journal of the Linnean Society.53 (2): 87–175.
Martin T et al. 2015. A Cretaceous eutriconodont and integument evolution of early mammals. Nature 526:380-384. online.

wiki/Eutriconodonta

Turtles as Hopeful Monsters – Future book and past essay by O. Rieppel

Figure 1. Turtles as Hopeful Monsters by O. Rieppel 2017.

Figure 1. Turtles as Hopeful Monsters by O. Rieppel, a book due to be published in 2017. The cover pictures Odontochelys, the earliest known soft shell turtle. The 2001 summary with the same title by O. Rieppel is the subject of the present blogpost.

Summary of this blogpost:
Without a large gamut phylogenetic analysis, such as the large reptile tree, that recovers two turtle clades derived from two phylogenetically reduced pareiasaur clades, any discussion of the origin of turtles is handicapped and will suffer from a surfeit of guesswork and error due to taxon exclusion. Sclerosaurus, Meiolania and Elginia are rarely considered in such studies, but are key to understanding turtle origins. Thankfully we have excellent embryological studies that more or less recapitulate phylogeny.

Dr. Rieppel has long been an advocate of a diapsid/placodont ancestry for turtles, and has applied molecule evidence to link turtles with archosaurs. He was part of the Odontochelys (Fig.1 ) team.

The hopeful monsters hypothesis is a biological theory which suggests that major evolutionary transformations have occurred in large leaps between species due to macro mutations. The LRT does not support major evolutionary transformations. All transitional taxa greatly resemble their nested sisters and microevolution is the only factor at play here.

From the Rieppel summary:
“A recently published study on the development of the turtle shell highlights the important role that development plays in the origin of evolutionary novelties.”

“Early theories attempted to explain the evolution of the turtle shell in the context of a step-wise, hence gradual process of transformation. The distant ancestor of turtles was hypothesized to have had a body loosely covered by osteoderms. Within the evolutionary lineage leading to turtles, the number of osteoderms would have gradually increased, until the bony plates would eventually have provided a complete covering of the trunk, thus forming an epitheca. Thecal ossifications would have developed below the epi- theca at later stages in the evolution of the turtle body plan, while epithecal ossifications would subsequently be lost at even more advanced stages of turtle evolution. This theory met with various difficulties, however, such as the fact that the earliest fossil turtle (Proganochelys) from the Upper Triassic of Europe (215 Mio years) has a complete theca. Furthermore, epithecal ossifications appear later than ossifications of the theca in development and, in modern turtles, epithecal ossifications tend to form in evolutionarily relatively advanced forms only.

“Turtles are unique among tetrapods, however, in that the shoulder blade (scapula) lies inside the rib cage (Fig. 3). The reason for this inverse relationship of the scapula is the close association of the ribs with the costal plates of the theca. The scapula of turtles comes to lie inside the rib cage because of a deflection of rib growth to a more superficial position. Recent developmental work has identified inductive interaction generated by the carapacial ridge as probable cause of this deflection of rib growth.

“A recent theory proposed the evolution of turtles from Paleozoic Pareiasaurs by a process of “correlated progression”. Correlated key elements of this progressive transformation are an increase in the number of osteoderms until they form a closed dorsal shield (carapace), the broadening of the ribs below this dorsal hield, the shortening of the trunk, the immobilization of the dorsal vertebral column and a backwards shift of the pectoral girdle.”

In the phylogenetic analysis provided by the LRT
turtles have two parallel origins, both from pareiasaurs: One for the domed, hard-shell clade (Bunostegos > Elginia > Meiolania) and one for the flattened soft-shell clade (Sclerosaurus > Odontochelys > Trionyx). In both these cases it appears that the tall scapula extended to either side of the narrow, cervical-like dorsal ribs. Then the anterior dorsal ribs rotated anteriorly over the tall scapulae, paralleling the rotation of the elbow anteriorly. You’ll note that turtles have more cervicals and fewer dorsals than pareiasaurs. The posterior cervicals in turtles appear to be the former dorsals of pareiasaurs. So, the pectoral girdle did not shift, but the posterior cervicals and anterior dorsals transformed around them. And this occurred during phylogenetic miniaturization.

Figure 3. Soft shell turtle evolution featuring Arganaceras, Sclerosaurus, Odontochelys and Trionyx.

Figure 3. Soft shell turtle evolution featuring Arganaceras, Sclerosaurus, Odontochelys and Trionyx.

Distinct from other reptiles
turtles do not have lateral movement in the torso with precursors in the short-torsoed, heavily ribbed pareiasaurs. “This repositioning of the vertebrae relative to the primary body segments is achieved by resegmentation of the somites. Each somite splits in half, and the posterior part of one somite recombines with the anterior part of the succeeding somite to form a vertebra,” reports Rieppel.

Figure 1. Hard shell turtle evolution with Bunostegos, Elginia, Meiolania and Proganochelys.

Figure 1. Hard shell turtle evolution with Bunostegos, Elginia, Meiolania and Proganochelys. The narrow and tall scapulae of pareiasaurs are carried forward in their descendant turtles during phylogenetic miniaturization. 

More from Dr. Rieppel
“As a turtle embryo grows and develops, the contours of the future carapace are soon mapped out by an accelerated growth and a thickening of the skin on its back.”

“The ribs of turtles are unique among vertebrates in that they chondrify within the deep layers of the thickened dermis of the carapacial disk.”

It has been known for over 100 years that in turtles, the neural arches of the dorsal vertebrae shift forward by half a segment, carrying the ribs with them, again a unique condition in amniotes.”

Embryological studies
(Ruckes 1929) indicate, “the scapula of turtles comes to lie inside the rib cage because of a deflection of rib growth to a more superficial position. The ribs come to lie lateral to the no longer functional intervertebral joints. The functional reason for this anterior shift of the neural arches is not clear, other than that it may contribute to the mechanical strength of the carapace, as the neural plates come to alternate with the costal plates. Neural and costal plates are endoskeletal components of the turtle carapace, and cannot be derived from a hypothetical ancestral condition by fusion of exoskeletal osteoderms. All other parts of the turtle carapace are exoskeletal.” Rieppel reports, “The turtle body plan is evidently highly derived, indeed unique among tetrapods.”

I’m going to say ‘not true’ here…
Several turtle-like forms developed among placodonts and the two turtle clades developed independently in parallel. Glyptodon had a turtle-like carapace, but no plastron. Even Minimi, the phytodinosaur, developed a club tail, convergent with meiolaniids.

Moreover, only the shells of the two clades of turtles are unique unto themselves, as most other of their body parts are microevolutionary adjustments from their separate micro pareiasaur bauplans. And, based on current fossil chronology, they had 25 to 45 million years to develop their respective shells from their proximal outgroup sisters.

References
Gilbert SF, Loredo GA, Brukman A, Burke AC. 2001.  Morphogenesis of the turtle shell: the development of a novel structure in tetrapod evolution. Evol Dev 2001;3:47 ± 58.
Götte A 1899. Über die Entwicklung des knoÈ chernen RuÈ ckenschildes (Carapax) der SchildkroÈten. Z wiss Zool 1899;66:407±434.
Lee MSY 1996. Correlated progression and the origin of turtles. Nature 1996;379:811- 815.
Rieppel O 1996. Turtles as diapsid reptiles. Nature 384 (6608), 453-455
Rieppel O 2001. Turtles as hopeful monsters. BioEssays 23:987-991.|
Rieppel O 2012. The Evolution of the Turtle Shell. Morphology and Evolution of Turtles. Part 2, 51-61. DOI: 10.1007/978-94-007-4309-0_5
Rieppel O 2017. Turtles as Hopeful Monsters. Origins and Evolution. Indiana University Press. 212 pp.. Online here.
Ruckes H. 1929 (12) Studies in chelonian osteology. Part II. The morphological relationships between girdles, ribs and carapace. Ann NY Acad Sci 1929;31:81-120.

Rieppel book summary online

Whales: Are they artiodactyls or tenrecs?

According to Wikipedia
Artiocetus (Gingerich et al. 2001; meaning essentially = artiodactyl + whale) was the first fossil to show that early whales possessed artiodactyl-like ankles (Fig. 1). The sizes, shapes and configurations are indeed similar. And this was a valid conclusion, but it was based on taxon exclusion. Other taxa have such ankles.

Figure 1. A selection of two whale ankles and one artiodactyl (Antilocapra, the pronghorn antelope). Such comparisons are the basis for aligning whales with artiodactyls.

Figure 1. A selection of two whale ankles and one artiodactyl (Antilocapra, the pronghorn antelope). Such comparisons are the basis for aligning whales with artiodactyls, but tenrecs were not considered.

Contra the artiodactyl hypothesis
the large reptile tree nests the basal whale, Maiacetus, with a clade of small and large, extinct and extant tenrecs like Hemicentetes, Andrewsarchus and Leptictidium.

So, to shed light on this disparity
here are two whale feet and ankles alongside the foot and ankle of Hemicentetes (Fig. 2), the tenrec without a tail (some other tenrecs have long tails (Figs. 5, 6), but extant examples don’t have online skeletons).

Figure 2. The evolution of the tenrec (Hemicentetes) pes, through the land whale Rhodhocetus and Basilosaurus.

Figure 2. The evolution of the tenrec (Hemicentetes) pes, through the land whale Rhodhocetus and Basilosaurus. Rhodocetus loses pedal digit 1. Basilosaurus loses pedal digit 2.  Note the lack of sharp claws and the lack of artiodactyl hooves in all taxa. Note the plantigrade pes here, not the digitigrade pes of artiodactylus. Note the reduction of distal tarsals (cuneiforms) down to dt3 (lateral cuneiform) in all taxa.

To the credit of the tenrec-whale clade:

  1. Hemicentetes has a short pedal digit 1. Rhodhocetus loses pedal digit 1. Basilosaurus loses pedal digit 2 (Fig. 2) where the vestigial pes is no longer in use.
  2. Tiny hooves on relatively slender digits
  3. Lack of tightly appressed artiodactyl metatarsals in all taxa.
  4. Plantigrade pes in all taxa
Figure 3. A basal artiodactyl, Ancodus, pes. As in Rhodhocetus pedal digit 1 is absent and the distal tarsals are reduced to one. This led to the artiodactyl hypothesis, and that is a great first guess! But it is not supported by the LRT.

Figure 3. A basal artiodactyl, Ancodus, pes. As in Rhodhocetus pedal digit 1 is absent and the distal tarsals are reduced to one. This led to the artiodactyl hypothesis, and that is a great first guess! But it is not supported by the LRT.

To the credit of the artiodactyl-whale clade hypothesis:

  1. Four pedal digits only, digit 1 is already absent
  2. The calcaneal heel is elongate.
  3. Metatarsals 2 and 5 axially rotate behind 3 and 4.
Figure 4. Maiacetus is a basal whale with legs and it is also a giant tenrec. Compare Leptictidium (Figs. 5, 6).

Figure 4. Maiacetus is a basal whale with legs and it is also a giant tenrec. Compare Leptictidium (Figs. 5, 6).

Further notes

  1. In both artiodactyls and tenrecs: reduction of distal tarsals (cuneiforms, yellow here) down to dt3 (the lateral cuneiform) in all taxa and a long list of other similar traits, both shared with proto-whales.
  2. Leptictidium (Fig. nests closer to whales and has a longer calcaneal heel with appressed metatarsals and reduced digits 1 and 5, but all toes are slender as in Rhodhocetus and it has a long muscular tail.

Artiodactyls came to the mind of Gingerich first
because artiodactyls had similar ankles to his land whale discoveries and were of similar size. Few workers both to study tenrecs — but tenrecs should have been included as they have similar pedes and taxon exclusion often arises with enigma taxa, as whales were.

  1. Artiodactyls are herbivores. Whales and tenrecs are not.
  2. Artiodactyls are digitigrade. Whales and tenrecs are not.
  3. Artiodactyls do not echolocate. Some tenrecs and some whales do.
  4. Artiodactyls do not have a large muscular tail for aquatic locomotion. Some tenrecs and all whales do.
  5. In the large reptile tree Maiacetus nests with a long list of tenrecs, not with artiodactyls, when given the opportunity, not with Ancodus.
  6. Several former mesonychids now nest with tenrecs and away from Mesonyx. clearing up that lingering issue.
  7. And a long list of traits in the skull and elsewhere…
Figure 2. Elements of Leptictidium from Storch and Lister 1985.

Figure 5. Elements of Leptictidium from Storch and Lister 1985. Note the long calcaneal heel here.

Added a few days later: here (Fig. 5a) is the double-pulley shape of the Leptictidium astragalus (n yellow).

Figure added later. The ankle of Leptictidium includes a spool-shaped double-pulley astragalus.

Figure 5a. The ankle of Leptictidium includes a spool-shaped double-pulley astragalus.

Figure 5. Leptictidium - Often considered a kangaroo like hopper (saltator), the loose sacral connection and phylogenetic bracketing suggest this was a dorsoventral undulatory swimmer instead. 

Figure 6. Leptictidium – Often considered a kangaroo like hopper (saltator), the loose sacral connection and phylogenetic bracketing suggest this was a dorsoventral undulatory swimmer instead.

Figure 6. Science magazine cover for Gingerich et al. 2001. Artist: John Klausmeyer.

Figure 7. Science magazine cover for Gingerich et al. 2001. Artist: John Klausmeyer. The hands and feet are far from being ungulate hooves, but close to tenrec paws.

Earlier notes
here, herehere and here. Unfortunately, as in several other taxonomic enigmas and mismatches, taxon exclusion prevented Gingerich et al. from making the tenrec connection.

References
Gingerich PD, Haq M, Zalmout I, Khan I and Malkani M 2001. Origin of whales from early artiodactyls: hands and feet of Eocene Protocetidae from Pakistan”. Science293 (5538): 2239–42.

Gingerich’s website
has more information. His Elomeryx is very close to Ancodus.

wiki/Artiocetus

Imagining the unknown: the skulls of Amphitherium and Docodon

Often enough
tiiny Jurassic synapsids, like Amphitherium prevosti  (von Meyer 1832; Middle Jurassic, 170 mya) and Docodon victor (Marsh 1881; Late Jurassic, 2 cm skull length), are known only from mandibles with teeth (Fig. 1).

We can guess what the skull looks like
because the molars occlude and the rest of the teeth interlock, slide past one another or meet at or near their tips. Plus we have clues from sister taxa that set parameters for possibilities in a method known as phylogenetic bracketing. In such cases some scores are less risky to guess, like the number of molars. Others are more risky, like the presence of caniniform canines.

Figure 1. Amphitherium and Docodon with skulls imagined.

Figure 1. Amphitherium and Docodon with skulls imagined. The large number of molars nests both these taxa with Monotremata.

References|
Marsh OC 1881. Notice of new Jurassic mammals: American Journal of Science, ser. 3, 21: p. 511-513.
Meyer H von 1832. Palaeologica, zur Geschichte der Erde und ihrer Geschöpfe. Schmerber, Frankfurt a/M, xi, 560 pp.