What is Periptychus carinidens?

Figure 1. Subset of the LRT focusing on the nesting of Periptychus.

Figure 1. Subset of the LRT focusing on the nesting of Periptychus.

Short answer:
In the large reptile tree (LRT, 1253 taxa, subset Fig. 1) Periptychus (Figs, 2–4) nests between basal phenacodonts like Phenacodus, Thomashuxleya and Pleuraspidotherium, and derived phenacodonts, like Gobiatherium + Arsinoitherium and Coryphodon + Uintatherium. These are all extinct herbivores from a clade that was recovered here first. Some derived taxa had ornate skull bumps/horns.

Previously known
as a condylarth from less complete materials (Cope 1881), the latest academic paper on Periptychus (Shelley, Williamson and Brusatte 2018) was still unable to determine closest relatives based on new data. No cladogram was presented. Sisters listed above were not listed in the text. Rather, the authors called it, “A robust, ungulate-like placental mammal.” 

Figure 2. Periptychus skull in 3 views.

Figure 2. Periptychus skull in 3 views ftom Shelley, Williamson and Brusatte 2018, colors added.

Think of Periptychus as a placental herbivore with very primitive feet…

Figure 3. Periptychus skeleton restored.

Figure 3. Periptychus skeleton restored from Shelley, Williamson and Brusatte 2018.

… and hands (no reduced digits). This mammal is remarkable for its long list of unremarkable traits.

Of course,
this was only the ‘warm-up act’ for the big, bizarre uintatheres to follow.

Figure 4. Manus and pes of Periptychus with some bones restored.

Figure 4. Manus and pes of Periptychus with some bones restored.

 

References
Cope ED 1881. The Condylarthra (Continued). American Naturalist 84;18: 892–906.
Shelley SL, Williamson TE and Brusatte SL 2018. The osteology of Periptychus carinidens: A robust, ungulate-like placental mammal (Mammalia: Periptychidae) from the Paleocene of North America. PLoS ONE 13(7): e0200132.

https://doi.org/10.1371/journal.pone.0200132

Advertisements

Let’s look at bandicoots!

Although bandicoots
(genus: Perameles, Fig. 1) are omnivores, they are basal taxa in the herbivorous metatherian clade in the large reptile tree (LRT, 1242 taxa, Fig. 4). They are closest to digging and burrowing marsupials, like the golden mole, Notoryctes, yet still close to the ancestry of the fast, leaping kangaroos, like Macropus. And they are derived from a sister to Arctocyon, the large carnivore.

Figure 1. Perameles nasuta, the long-nosed bandicoot, nests as basal member of the herbivorous marsupials.

Figure 1. Perameles nasuta, the long-nosed bandicoot, nests as basal member of the herbivorous marsupials. Look at those huge lumbar vertebrae! Scale bar = 1cm.

Perameles nasuta (Geoffroy 1804, extant, 40 cm long) is the long nosed bandicoot. This basal marsupial is a nocturnal omnivore, like its ancestor Didelphis (Fig. 2). It has a rear-facing pouch, three long digging fingers, two short ones and a pes dominated by pedal digit 4. Digits 2 and 3 were reduced to grooming claws, as in kangaroos. Gestation lasts a mere 12.5 days, then the young spend another 8 weeks in the marsupium. The long nose of bandicoots is a basal trait retained by Ukhaatherium and prototheres like Cronopio and Juramaia.

YouTube videos
of bandicoots show they have a high metabolism and can scoot away rapidly, like rats or rabbits, distinct from their slower moving opossum ancestors. Of course, Perameles was ancestral to kangaroos, like Macropus (Fig. 3), so their leaping ability is nascent here and their odd feet are nearly identical. Kangaroo hands are still primitive with five sub-equal fingers, not evolved for digging, like odd hand of Perameles (Fig. 1).

According to Wikipedia
“The position of the Peramelemorphia within the marsupial family tree has long been puzzling and controversial. There are two morphological features in the order that appear to show a clear evolutionary link with another marsupial group: the type of foot, and the teeth. Unfortunately, these clear signposts point in opposite directions.” The LRT solves this problem by nesting omnivorous Perameles at the base of the herbivores and allowing for convergence between the large kangaroo and wombat anterior dentaries.

Adding taxa
is helping to clarify phylogenetic relationships among the marsupials. We’ll look at these soon.

References
Geoffroy E. 1804. Mémoire sur un nouveau genre de mammifères á bourse, nommé Perameles. Annales de la Musee National d’ Histoire Naturelle de Paris 4: 56–64.

Armadillosuchus: another small herbivorous croc

Marinho and Carvalho 2009
brought us a new, small, crocodyliform from the Late Cretaceous of South America, Armadillosuchus (Fig. 1).

Oddly,
the nares are not mentioned in the text, nor labeled in the published figure (Fig. 1). Where are they?

Figure 1. Armadillosuchus skull in dorsal view and lateral view. A second specimen preserves teeth. Reconstruction below aligns the ventral maxilla with the quadrate, as in all sister taxa.

Figure 1. Armadillosuchus skull in dorsal view and lateral view. A second specimen preserves teeth. Reconstruction below aligns the ventral maxilla with the quadrate, as in all sister taxa. Where are the nares? They are not mentioned in the text.

Armadillosuchus arrudai (Marinho and Carvalho 2009; Late Cretaceous; est. 2m in length) was an herbivorous and armored crocodylomorph from South America. It nests with Mariliasuchus (above) in the LRT. Both have giant premaxillary teeth. The naris is reduced to a tiny hole facing anteriorly. The rostrum may have been more horizontal than originally reconstructed as all sister taxa line up the quadrate with the ventral maxilla (Fig. 1 bottom figure). These are members of the Ziphosuchia.

Figure 1. Mariliasuchus skull in several views. Note the premaxillaery fangs and the short blunt remainder of the teeth.

Figure 2. Mariliasuchus skull in several views. Note the premaxillaery fangs and the short blunt remainder of the teeth. The nares are very tiny and anteriorly oriented. Note the alignment of the quadrate with the ventral rim of the maxilla together with the rostrum vs. forehead angle, as in the new lateral view of Armadillosuchus (Fig. 1).

The nares of Armadillosuchus
are best found by phylogenetic bracketing based on the nares found in its sister taxon Mariliasuchus (Fig. 2).

Figure 2. Subset of the LRT focusing on Crocodylomorpha (basal Archosauria) including Armadillosuchus.

Figure 2. Subset of the LRT focusing on Crocodylomorpha (basal Archosauria) including Armadillosuchus.

Key to understanding the origin of the clade Dinosauria
is to understand the proximal outgroup taxa, the bipedal basal Crocodylomorpha, which no prior studies include (though some include Lewisuchus).

Images of complete skeletons of Armadillosuchus
are online, photographed from museum mounts. I have not found academic data for anything more than is figured above. The rest may be restored. Let me know of any citations I have missed.

References
Marinho T S and Carvalho, IS 2009. An armadillo-like sphagesaurid crocodyliform from the Late Cretaceous of Brazil. Journal of South American Earth Sciences. 27 (1): 36–41.

wiki/Mariliasuchus
wiki/Armadillosuchus

The Captorhinidae: herbivory and rates of evolutionary change

Dr. Neil Brocklehurst brings new insight to herbivory and evolution as he
compares rates of evolution as reptiles venture into a previously unexploited diet: plants. I did not comment on PeerJ where it is currently published without peer review because I thought it would be better here and Dr. Broklehurst reads this blog.

From the Brocklehurst abstract:
“Here I examine the impact of diet evolution on rates of morphological change in one of the earliest tetrapod clades to evolve high-fibre herbivory: Captorhinidae. Using a method of calculating heterogeneity in rates of discrete character change across a phylogeny, it is shown that a significant increase in rates of evolution coincides with the transition to herbivory in captorhinids.”

FIgure 1. Subset of the LRT focusing on the Captorhinidae.

FIgure 1. Subset of the LRT focusing on the Captorhinidae. all herbivores.

Brocklehurst notes
“By the end of the Cisuralian (Early Permian), five tetrapod lineages had independently evolved a herbivorous diet (referencing Sues and Riesz 1998).”

  1. Captorhinidae
  2. Diadectidae
  3. Pareiasauridae
  4. Caseidae
  5. Edaphosauridae

Matching the Brocklehurst cladograms
In the LRT the basal herbivore is also Thuringothyris, and it nests close to the base of the new Lepidosauromorpha (Fig. 1) at the base of the Captorhinidae. One wonders if the original dichotomy of reptiles actually separated slightly larger herbivores from slightly smaller insectivores in the Viséan (Early Carboniferous)?  At present evidence only supports a later adoption of herbivory in the Late Carboniferous among several lepidosauromorph taxa. So there had to have been an earlier undiscovered origin. In any case the first four clades in the Sues and Riesz list (above green) are all related to each other in the clade Lepidosauromorpha. They all had a single ancestor (see below). Later lepidosauromorphs, like turtles, lizards, snakes and pterosaurs reacquired insectivory, piscivory and carnivory independently.

Urumqia liudaowanensis (Zhang et al. 1984) ~20 cm snout-vent length, Lower Permian.

Figure 3. Urumqia liudaowanensis (Zhang et al. 1984) ~20 cm snout-vent length, Lower Permian.

Late survivors of an earlier radiation
Urumqia (Fig. 3) nests as the basalmost lepidosauromorph, but fossils have only been found in Late Permian strata. Thus, Urumqia was a living fossil in the late Permian. Notably the gastralia were much wider than the posterior dorsal ribs. This created a large gut, ideal for herbivory (see below), but it also provided a larger volume for greater egg production.  Bruktererpeton was a sister and a basal lepidosauromorph with fossils found in Late Carboniferous strata with no obvious herbivorous traits. However, it too, nested with Thuringothyris (Fig. 1), so could have been an herbivore.

Figure 2. Captorhinidae according to Brockelhurst on PeerJ 2017. Most of the taxa also appear on the LRT, which is great case of congruence!

Figure 2. Captorhinidae according to Brocklehurst on PeerJ 2017. Most of the taxa also appear on the LRT, which is great case of congruence!

The taxon list
(Fig. 2) of Brocklehurst 2017 was restricted to his list of Captorhinidae. The LRT (Fig. 1) also nests most of his taxa within a single clade. However, Thuringothyris nests outside the Captorhinidae in the LRT but at its base. Saurorictus nests as the basal captorhinid in the LRT, despite its late appearance in the fossil record. It shares many traits with Millerettidae, a more derived taxon leading to all later lepidosauromorphs. Opisthodontosaurus appears in both cladograms, but its sister, Cephalerpeton appears only in the LRT. I have not yet seen data on Rhiodenticulatus and the derived captorhinid taxa are not present in the LRT. Limnoscelis and Orobates also nest as sisters to Saurorictus in the LRT. Limnoscelis is traditionally considered a carnivore, but since it is phylogenetically bracketed by herbivores, that hypothesis should be reexamined.

Sues and Reisz 1998 note:
“Dental features indicative of herbivorous habits include the presence of crushing and grinding dentitions, or marginal teeth with leaf-shaped, cuspidate crowns suitable for puncturing and shredding. Cranial features include short tooth rows and elevation or depression of the jaw joint for increased mechanical advantage during biting, large adductor chambers and deep lower jaws for accommodating large adductor jaw muscles, and jaw joints that permit fore-and-aft motion of the mandible.”

“The discovery of gut contents composed of conifer and pteridosperm ovules in specimens of the Late Permian diapsid reptile Protorosaurus (Munk and Sues 1992), long thought to be carnivorous based on its dentition, demonstrates that the consumption of plant material is not necessarily reflected by dental specialization.”

“The rib-cages of Palaeozoic herbivores are typically significantly wider and more capacious than those of their closest faunivorous relatives.”

Brocklehurst discusses rate variation:
“Discrete morphological character scores may be taken from the matrices used in cladistic analyses, and ancestral states are deduced using likelihood. This allows the number of character changes along each branch to be counted, and rates of character change are calculated by dividing the number of changes along a branch by the branch length. The absolute value calculated for the rate of each branch, however, can be misleading due to the presence of missing data (Lloyd et al. 2012). As such it is more useful to identify branches and clades where the rates of character change are significantly higher or lower than others, rather than comparing the raw numbers.”

Brocklehurst concludes:
“the evidence supporting an adaptive radiation of captorhinids coinciding with the origin of herbivory in this clade is compelling. It is only along herbivorous branches that significant increases in rates of morphological evolution are identified in the majority of the 100 time-calibrated trees.”

Brocklehurst has a good hypothesis with broader implications:
Among mammals, with the exception of tenrecs that turned into odontocete whales, the carnivores are more conservative than the herbivores, which developed horns, trunks and antlers, along with a variety of tooth morphologies. The clade Carnivora is quite conservative.

Among dinosaurs, with the exception of birds, the theropods are more conservative that the herbivores, which developed horns, long necks, great size, frills and duckbills.

Among basal reptiles, the lepidosauromorph herbivores developed into a wider variety of shapes and sizes while the archosauromorph insectivores were more conservative and stayed small until the advent of the lateral temporal fenestra that appeared in basal synapsids and diapsids.

References
Brocklehurst N 2017. Rates of morphological evolution in Captorhinidae: an adaptive radiation of Permian herbivores PeerJ Preprints (not peer-reviewed) PDF
Munk W and Sues H-D 1992. Gut contents of Parasaurus (Pareiasauria) and Protorosaurus (Archosauromorpha) from the Kupferschiefer (Upper Permian) of Hessen, Germany, Paläont. Z. 67, 169–176.
Sues H-D and Reisz RR 1998. Origins and early evolution of herbivory in tetrapods. Trends in Ecology and Evolution 13:141-145.

According to Wikipedia, PeerJ is 
“an open access peer-reviewed scientific mega journal covering research in the biological and medical sciences. PeerJ uses a business model that differs both from traditional publishers – in that no subscription fees are charged to its readers – and from the major open-access publishers in that the publication fees are levied not per article but per publishing researcher and at a much lower level. PeerJ charges authors a one-time membership fee that allows them – with some additional requirements, such as commenting upon, or reviewing, at least one paper per year – to publish in the journal for the rest of their life.[12] Submitted research is judged solely on scientific and methodological soundness (like at PLoS ONE), with peer reviews published alongside the papers.”

Maybe Paraceratherium is really a giant horse.

Figure 1. Subset of the large reptile tree focusing on horses and their kin.

Figure 1. Subset of the large reptile tree focusing on ungulates and their kin.

Today’s heresy began when several ungulate taxa
were added to the large reptile tree (LRT, Fig. 1, now 907 taxa, completely resolved with 228 traits). Equus, the horse; Paraceratherium, the giant hornless ‘rhino’; Ceratotherium the white rhinoceros and Embolotherium, a  Mongolian brontothere.

It is widely accepted
and supported by the LRT that horses and rhinos share a common ancestor (Fig. 1). In this case, the LRT recovered either 1) an overlooked relationship, or 2) a case of convergence between Paraceratherium and Equus (Fig. 2). Wikipedia  notes that rhinos are more closely related to tapirs. In the LRT tapirs are basal to a sister clade to the rhino/horse clade. Hyracotherium and Heptodon are basal to the rhino/horse clade.

Figure 1. Equus and Paraceratherium nest together on the LRT.

Figure 2. Equus and Paraceratherium currently nest together on the LRT. Additional taxa will, no doubt, change that, but at present, Paraceratherium shares more traits with Equus than Ceratotherium, the white rhino (Fig. 2). The long neck and premaxillay teeth, along with other traits, separate these two from extant rhinos.

Equus ferus (Linneaus 1758; Figs. 2,3) includes several extant horses, mules and zebras. Compared to Hyracotherium the Equus preorbital region is longer, the frontal produces a postorbital bar that contacts the squamosal. The premolars are molarized. All four limbs end in a single toe, digit 3.

Figure 1. Equus the extant horse.

Figure 3. Equus the extant horse has a postorbital bar, an elongate rostrum and a single toe on each limb.

Hyracotherium leporinum (Owen 1841; 78 cm long; Eocene, 55-45 mya; BMNH C21361nests basal to the horse/rhino clade, EquusHeptodon (Fig. 5) is an Eocene sister. Canines remained. A diastema separated the canine from the premolars. The manus has four hoofed toes. The pes has three hoofed toes. The premolars were becoming molarized. Compare the skull of dog-size Hyracotherium to the similar skull of Paraceratherium, one of the largest land mammals of all time and Equus, the horse.

Figure 2. Hyracotherium is an Eocene horse sister in the LRT. Skull bones are colorized here.

Figure 4. Hyracotherium is an Eocene horse sister in the LRT. Skull bones are colorized here.

Heptodon (Pachynolophus) posticus (Cope 1882; Eocene, 50 mya; 1m in length) was derived from a sister to Tapirus and was itself a sister to Hyracotherium and the base of brototheres like Embolotherium (below). Heptadon had reduced canines and developed a diastema (lack of teeth) posterior to them. The posterior cranium was slightly elevated. The manus had four digits. The pes had three. 

Figure 4. Heptodon originally nested with Tapirus, but with the addition of Equus, Hyracotherium and Embolotherium it shifted to nest with Embolotherium.

Figure 5. Heptodon earlier nested with Tapirus, but with the addition of Equus, Hyracotherium, Paraceratherium and Embolotherium it nested closer to them.

The extant white rhinoceros
Ceratotherium simus (extant, Figs. 6, 7)) shares little in common with Paraceratherium, but shares a long list of traits with Embolotherium, including an oddly elevated pair of nasals and the near complete loss of the lumbar region. Both had smaller ancestors, so direct comparisons yield several possible convergences that currently nest as homologs. That’s what happens with taxon exclusion.

Figure 6. Ceratotherium, simum, the white rhinoceros with keratinous horns in dark brown. Note the elevated nasals and convex dentary ventral margin.

Figure 6. Ceratotherium, simum, the white rhinoceros with keratinous horns in dark brown. Note the elevated nasals and convex dentary ventral margin.

Figure 7. Ceratotherium (white rhino) skeleton, distinct from the long-legged Paraceratherium.

Figure 7. Ceratotherium (white rhino) skeleton, distinct from the long-legged Paraceratherium.

The last taxon to be added is
Embolotherium andrewsi (Osborn 1929; Late Eocene; 2.5m tall at the shoulder; Mongolia; Figs. 4, 5). It nests as a sister to the rhino, Ceratotherium, but this is likely to be overturned on convergence when additional taxa are added. This highly derived brontothere (= titanothere) has forked ‘horns’ (= rams). The rams are elevated nasal bones, hollow and fragile. These are in contrast to the solid rams of North American brontotheres. Click here to see the original AMNH illustration of the Embolotherium skull that portrays the area beneath the elevated nasal as a thick fleshy area. Alternatively Wikipedia reports, “the bony nasal cavity extends to the peak of the ram, thus implying that the nasal chamber was greatly elevated, possibly creating a resonating chamber.”

Figure 2. Brontotherium, a sister to Embolotherium.

Figure 8. Brontotherium, a sister to Embolotherium, which is known from skulls, but no complete post-crania.

The skull of Embolotherium was 2x wider than tall at the orbits. The molars were much larger than the premolars, which were themselves molarized. The canines were vestiges. The posterior skull was greatly elevated. The dorsal spines were greatly elevated (Fig. 4). The lumbar region was reduced. The ilia were transverse. Four fingers are retained by the manus, indicating an early divergence from three-fingered horses and rhinos.

Figure 4. Embolotherium andrewsi modified from the AMNH website to show a possible inflatable narial area.

Figure 9. Embolotherium andrewsi animation modified from the AMNH website (see link above) to show a possible inflatable narial area, contra the original restoration with a fleshy, immobile sub-nasal area.

Figure 8. Paraceratherium pes. Note the reduced lateral and medial toes (2 and 4). As in Equus, and distinct from Ceratotherium, the central toe is much larger.

Figure 10. Paraceratherium pes. Note the reduced lateral and medial toes (2 and 4). As in Equus, and distinct from Ceratotherium, the central toe is much larger.

These are all perissodactyls
(odd-toed ungulates)
despite the fact that the manus has four digits. The pes (with toes) has an odd-number of digits (three or one). Wikipedia nests Desmostylia and Anthracobunidae among the perissodactyls, but they nest with hippos and mesonychids in the LRT.

Hyracodon
is not included here, but nests in the LRT between Hyracotherium and Heptodon, far from Paraceratherium. Apparently Equus and Paraceratherium have not been tested together in phylogenetic analysis under the assumption that one was a horse and the other a rhino.

Finally
note the large third digit of the pes of the giant three-toed horse, Paraceratherium (Fig. 10), as in Equus and distinct from Ceratotherium, which has three toes, but sub equal in size.

The above was dashed off
a little more quickly than usual. Please bring to my attention any typos or dangling hypotheses.

 

References
Cope ED 1882. Paleontological Bulletin 34:187.
Froehlich DJ 2002. Quo vadis eohippus? The systematics and taxonomy of the early Eocene equids (Perissodactyla). Zoological Journal of the Linnean Society. 134 (2): 141–256.
Linnaeus C 1758. Systema naturæ per regna tria naturæ, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Tomus I. Editio decima, reformata.
Osborn HF 1929. Embolotherium, gen. nov., of the Ulan Gochu, Mongolia. American Museum novitates; no. 353.
Owen R 1841. Description of the Fossil Remains of a Mammal (Hyracotherium leporinum) and of a Bird (Lithornis vulturinus) from the London Clay. Transactions of the Geological Society of London, Series 2, VI: 203-208.

wiki/Equus
wiki/Hyracotherium
wiki/Embolotherium
wiki/Paraceratherium
wiki/Heptodon

Titanoides: a rarely studied, but key mammal

Titanoides is rarely studied.
And it (Fig. 1) nests at the base of all large herbivorous mammals in the LRT (Fig. 2). That makes it more interesting than just another pantodont with a much smaller tail and larger canines than its phylogenetic sister, Barylambda nesting at the base of the Xenarathra (Fig. 2).

Figure 1. Titanoides nests between Barylambda (basal Xenarthra) and higher mammalian herbivores.

Figure 1. Titanoides nests between Barylambda (basal Xenarthra) and higher mammalian herbivores.

Titanoides primaevus (Gidley 1917; Late Paleocene, 58 mya; 3m in length) was originally considered a type of brontothere, but is now traditionally considered a pantodont mammal. Sabertooth canines mark this otherwise bear-like and bear-sized herbivore. Distinct from sister taxa, it had clawed digits. The skull was more than twice as wide as tall. Three premolars looked very much like molars. The jugal/zygomatic arch was slender and weak. The tail was a short vestige. Take a look at that odd lower canine (Fig. 1 orange tooth) with a concavity to accommodate the upper canine.

Figure 3. Subset of the LRT focusing on mammals. Titanoides nests at the base of all large herbivores.

Figure 3. Subset of the LRT focusing on mammals. Titanoides nests at the base of all large herbivores.

Sharp-eyed observers
will see a few very minor taxon shifts in the LRT (Fig. 2). This is what happens when more taxa are added. I also reexamined several poor scores and found a few errors that further cemented and slightly shifted a few relationships.

Happy holidays, everyone!
It’s gratifying to see a growing list of subscribers and readers. I wish you all the best!

References
Gidley JW 1917. Notice of new Paleocene mammal, a possible relative of the Titanotheres. Proceedings of the United States National Museum 52(2187):431-435.

wiki/Titanoides

Pleuraspidotherium and Orthaspidotherium

These two taxa don’t make very many lists.
That may be because Pleuraspidotherium amonieri  (Paleocene; Lemoine 1882; Fig. 1) and Orthaspidotherium edwardsi (Ladevèze, Missiaen and Smith 2010; ) are not directly related to any ‘big name’ clades and they have a very basal condylarth (herbivorous mammal) look. Instead these two nest with rather plesiomorphic Meniscotherium and highly derived Astrapotherium in the large reptile tree (LRT, 898 taxa). Both Pleuraspidotherium and Orthaspidotherium were originally recognized as phenacodontids related to Meniscotherium, so we’re tracking traditional nestings here.

Figure 1. Orthaspidotherium from x et al. 2009 is a plesiomorphic mammalian herbivore, basal to all later forms, from elephants to baleen whales to giraffes.

Figure 1. Orthaspidotherium from Ladevèze, Missiaen and Smith 2010 is a plesiomorphic mammalian herbivore, basal to all later forms, from elephants to baleen whales to giraffes.

However in the LRT,
both taxa nest together and apart from Phenacodus.

Figure 1. Pleurospidotherium a

Figure 1. Pleuraspidotherium a

These two were among the very first
slightly larger mammalian herbivores that first appeared in the Cenozoic. We see the origin of the notched diastema here separating the anterior premolars from the poster premolars and molars.

Figure 2. Subset of the LRT showing the nestings of Pleurospidotherium and Orthaspidotherium at the base of the herbivorous mammals.

Figure 2. Subset of the LRT showing the nestings of Pleurospidotherium and Orthaspidotherium at the base of the herbivorous mammals.

Halliday 2015 reports
“Ladevéze et al. (2010) hypothesised that Pleuraspidotheriidae are closest relatives to arctocyonids such as Chriacus, in a group also including the basal artiodactyls, but their taxonomic sampling was very low, and only very few representatives of each supposed group were present.” 

In the LRT
Chriacus nests with bats and Arctocyon nests with Didelphis in the Metatheria, both far from the Pleuraspidotheriidae. None of these relationships is found in Halliday et al. 2015.

Halliday 2015 reports, 
“With the exception of Primates (Russell, 1964), Rodentia (Jepsen, 1937), and Carnivora (Fox, Scott & Rankin, 2010), no extant order of placental mammal has an unambiguous representative during the Paleocene.” Pleuraspidotherium and Orthaspidotherium are also in the Paleocene, so they are early representatives of the herbivorous placental clades.

“Despite numerous suggestions of Cretaceous placentals, no Cretaceous eutherian mammal has been unambiguously resolved within the placental crown.” In the LRT multituberculates and Shenshou from the Jurassic are rodent sisters, Volaticotherium is a basal pre-placental from the earliest Cretaceous. Docofossor is a basal Oxfordian (early Late Jurassic) marsupial.  Maotherium a pre-mammal from the early Cretaceous, Zhangheotherium, a basal pangolin is from the earliest Cretaceous, and Maelestes, a basal tenrec is from the late Cretaceous, so the Halliday claim is not validated by the LRT and a Cretaceous origin would therefore NOT require the existence of long ghost lineages, contra Halliday et al. 2015.

Halliday et al. 2015 illustrates
the ‘current consensus” of mammalian relationships with the first split at Xenarthra + Tenrecoidea and kin splitting from Glires + the rest of the placentals in something of a mishmash of tree branches. The LRT, by contrast, recovers complete resolution at all branches and does not replicate the “consensus” topology.

Halliday et al. then reports on their own phylogenetic analysis based on 680 traits and 177 taxa. The resulting topology bears little similarity to the the LRT with the first split separating (primates + plesiadapids) + (rodents + rabbits) + xenarthra  from the rest of the placentals, then Phenacodus + Meniscotherium and kin splitting next from the remaining placentals in one test.

Another result split Xenarthra and Procavia + Potamogale and kin from the rest of the mammals. Among their seven conclusions, they report, “No definitive crown-placental mammal has yet been found from the Cretaceous, as Protungulatum is resolved as a stem eutherian, and therefore the Cretaceous occurrence of Protungulatum cannot be considered definitive proof of a Cretaceous origin for placental mammals.”

This is contradicted by the LRT results.

References
Halliday T et al. 2015. Resolving the relationships of Paleocene placental mammals. Biologoical Reviews. doi: 10.1111/brv.12242
Ladevèze S, Missiaen P and Smith T 2010. First Skull Of Orthaspidotherium edwardsi (Mammalia, “Condylarthra”) From The Late Paleocene Of Berru (France) And Phylogenetic Affinities Of The Enigmatic European Family Pleuraspidotheriidae”. Journal of Vertebrate Paleontology. 30 (5): 1559–1578.
Lemoine V 1882. Sure l’encephale de l’Artocyon et du Pleurospidotherium aumonieri. Bulletin de la Societe Géologie de France 3 series t. X. Also. Comptes Rendus.

wiki/Orthaspidotherium
wiki/Pleuraspidotherium