Re: *Silesaurus*: the basalmost ornithischian after all?

["Jaime Headden" <ja_headden@qilong.8m.com>]

David Marjanovic (david.marjanovic@gmx.at) wrote:

<In its description, *Silesaurus* is provisionally excluded from Dinosauria on 
the basis of
 - its very proximally located deltopectoral crest (but it is long enough for a 
dinosaur)
 - and its lack of epapophyses (which according to Fraser et al. [2002, 
description of *Agnosphitys*] are only known in
"herrerasaurs and theropods" -- did they overlook...?).>

  The actual spelling of the name is *Agnostiphys,* not *Agnosphitys;* check 
comments online, from Darren Naish and others.

<But it shares with ornithischians, to the exclusion of (basal) saurischians, 
the following features (quotes from Glut's
encyclopedia, p. 62):

  - herbivory, with assorted dental characters:
    - "cheek teeth with low triangular crowns with well-developed cingulum 
beneath"
    - "cheek teeth with crowns having low and bulbous base"
    - "maxillary and dentary teeth with overlapping adjacent crowns" -- to a 
very small extent in the mx, somewhat more in the d
    - "maxillary and dentary teeth not recurved" -- well, they are, but very 
little"
    - "maximum tooth size near middle of maxillary and dentary tooth rows">

  "Herbivory" should, expectantly, form the same or similar complex of 
anatomical features in a jaw among dissimilar animals.
Herbivorous reptiles are notorious for developing features similar to one 
another, as advanced aetosaur teeth are superfically
similar to teeth of ankylosaurs or stegosaurs, and derived notosuchians have 
multicuspidate, multi-rowed molariform teeth, virtually
identical to traversodont cynodontian and multituberculate molars. Herbivory, 
unlike carnivory, enforces that the bite will be
stronger closer to the origin of the adductors, so that teeth towards the rear 
of the jaw are larger than towards the tip, usually
the reverse in carnivores, and should be expected in any herbivorous amniote, 
as even several lacertilians parallel herbivorous
dinosaurs in having multicuspidate teeth arranged _en echelon_; one reason why 
teeth in troodontids are arranged the way they are is
likely to enforce an herbivorous shear mechanism in the front of the snout 
without sacrificing the teeth themselves (the
shear-and-tear mechanic which is teeth in several ornithischians with 
premaxillary teeth). Shapes of the teeth in several basal
dinosaurs, including *Saturnalia,* *Eoraptor,* and *Lesothosaurus* are 
identical, with only slight variations; for instance,
*Lesothosaurus* lacks distint cingula entirely, and *Saturnalia* has slight 
cingula as well, whereas half the teeth in *Eoraptor*
are virtually phyllodont with expanded bases, some of which with low-aspect 
crowns. According to the bones illustrated in Dzik
(2003), the crowns are not arranged _en echelon_, and are well-defined from 
one-another, so that Glut's quote, presumably from the
3rd supplement, doesn't seem at all supported. The teeth are conical, slightly 
compressed linguolabially (mediolaterally) and only
slightly recurved; some have an aspect to teeth in *Thecodontosaurus* and 
*Saturnalia* in being high-aspect and triangular with a
straight distal (caudal) margin, and a curved mesial (rostral) one; 
longitudinal ridges are not radiate, but convergent to the apex,
as in crocs, mosasaurs, labyrinthodont amphibians, spinosaurs, and most other 
piscivorous tetrapods and some fish.

  This complex of features are thus not consistently arguable for relationship. 
Similarly, they argue that a dental suite among
basal dinosaurs were not herbivorous (absence of a distincty phyllodonty, for 
examply) but are likely omnivorous instead. The dental
anatomy of *Silesaurus* lacks any real herbivrous adaptations, contra Dzik, in 
that they are carinate, conical, variably and very
briefly denticulated, with most teeth having distinct recurvature throughout 
the series, and thus resemble nothing even so much as
the most basal ornithischians, much less the dental anatomy of derived 
ornithischians.

<  - a beak on the tip of the lower jaw. There is no predentary, but because 
the beak must have come before the predentary (assuming
that the predentary, like the rostral, evolved to give additional support to 
the beak), this should not exclude *Silesaurus* from
being the basalmost ornithischian.>

  This doesn't really work as most work appears to support that a keratinous 
beak it not neccessary for either herbivory, or formed
prior to a toothless jaw. It seems likely, but there has yet to be any proof 
for this. The lower jaw's "hook" apparently interacted
only with teeth, as it is narrower than the paired premaxillary dentition, so 
the possibility of a lower "beak" to propose the
presence of an upper beak, as in ornithischians, co-existent with a reduction 
in dentition relative to premaxillary length, is not
viable. The "hook" therefore may have served a separate function than cropping 
vegetation; the paired dentaries would form a apex in
cross-section higher medially than laterally, and is remarkably unlike that of 
the predentary morphology of basal ornithischians,
and I would suggest that it not be considered homologous to the predentary of 
ornithischians.

< - 4 sacrals (respectively 4 fused vertebrae, 2 of which contact the ilia). 
Ornithischians start at 5 (all of which contact the
ilia), basal saurischians retain 2.>

  This doesn't really support anything. *Eoraptor,* as stated elsewhere, has 
_three_ not _two_ sacrals. Sacral count is also
variable by age, where count increases in *Coelophysis,* prosauropods, and 
*Silesaurus,* where almost similarly-sized individuals
have a loose caudosacral, dorsosacral, or unfused sacrum entirely. The 
condition in *Silesaurus* appears to be unique, and
age-dependant.

< - the lack of the 5th toe.>

  Pardon? The fifth metatarsal in some saurischians has a phalanx on it, but 
this is hardly a toe. However, because the foot of
*Silesaurus* is so derived, with a fifth metatarsal that is shortened and 
semilunate as in tetanuran theropods only, this is
unlikely to be a condition relating to ornithischians, especially as the foot 
has other advanced, cursorial adaptation (short toes,
shortened second and fourth metatarsals relative to the third) that is unlike 
anything but a theropod.

< - a (slight) crossing of radius and ulna>

  Highly plesiomorphic, as both ornithischians and saurischians have crossed 
ulnae. I fthis refers to the relatively non-crossed
ulnae, this is inherent in quadrupeds as much as it is advanced in bipeds due 
to restriction of pronatory ability of the wrist; the
long forelimbs thus argue for quadrupedal adaptations and, like the (obviously 
non-ornithischian *Marasuchus* which has similar
limbs, or *Lewisuchus* with similar humerus, or *Scleromochlus* with similar 
limb structures as well).

< - reduced postcranial pneumaticity -- at least that's how the "chonoi" of 
*Silesaurus*, which fail to enter the vertebrae, and the
absence of pneumatic features in undoubted ornithischians can be interpreted.>

  "Reduced" argues there was some to begin with, and the condition was 
reversed. "Limited" or "incipient" argues it is present only
in minor condition and increased. However, if the argument is due to vertebral 
"chonoi" or the "chonoses" between successive
vertebrae, or the fossae between the centrodiapophyseal laminae of the the 
vertebrae themselves, these are also present in
non-dinosaurs like *Marasuchus* and *Lewisuchus* and are limited in 
*Herrerasaurus* and other basal dinosaurs, implying this is a
plesiomorphy, and is "incipient," rather than "reduced." Neurocentral or 
central fossae do not occur in basal theropods, and are
unexpected to occur.

<  - "large lateral process of premaxilla excluding maxilla from margin of 
external naris". I don't know how large large is, but,
judging from the excavations on the nasal and mx, there was such an unpreserved 
process. Even though ns and mx still met underneath
it.>

  A ventral process of the nasal separates the maxilla from the tip of the 
caudoventral process of the premaxilla, interfingering
between the premaxilla and maxilla. This does not otherwise occur in basal 
dinosaurs, saurischian or ornithischian, and has no
bearing on its relationship with regards to ornithischians.

<  - "prefrontal with long caudal ramus overlapping frontal". Again not 
quantified, but, judging from the facet on the frontal, the
unpreserved prf did overlap it for over 1/5 of the latter's length. But I 
wonder if that isn't a plesiomorphy. It is a synapomorphy
of Dinosauria (or something more inclusive) that the frontal participates at 
all in the margin of the orbit, right?>

  for the most part, apparently. The skull of *Marasuchus* is not as well 
preserved as that of *Lewisuchus,* but it seems that in
either, there is an orbital rim for the frontal. A long articulation for the 
prefrontal/lachyrmal occurs in *Herrerasaurus* as well
as "prosauropods" and *Eoraptor*, and to a more limited extent in basal 
ornithischians such as *Lesothosaurus.* Thus, this feature,
also does not corroborate ornithischian features of *Silesaurus.*

<  - "palatal process of premaxilla horizontal or broadly arched" -- not quite 
clear but likely from fig. 6.>

  The medial contact between both maxilla is a symptom of a continuous, and 
convergent, secondary palate that extends below the
external nares and further caudally. This does not otherwise occur in 
ornithischians. The horizontal or "arched" condition occurs in
theropods and "prosauropods" as well.

<  - "quadrate massive, elongate" -- yes, if I've correctly interpreted the 
absent quantification. Looks pretty standard to me, like
in theropods. Hm.>

  Actually, the condition resembles *Lewisuchus,* "prosauropods" and theropods 
quite well, whereas in ornithischians the quadrate is
nearly straight on the caudal margin with a caudally inflected proximal 
(sphenoid condyle) end. Otherwise, it appears plesiomorphic
with respect to basal dinosaurs.

<  - "lateral swelling of ischial tuberosity of ilium"; is this the 
antitrochanter, which according to Fraser et al. is a much more
widespread feature?>

  Most likely this is the antitrochanteric crest, which causes the ischiadic 
peduncle to be triangular in cross-section along a
horizontal plane. Seen in *Caseosaurus,* *Herrerasaurus,* *Lesothosaurus,* and 
*Guaibasaurus.*

<  - absence of gastralia, assuming we can trust the fact that none are 
preserved. But clavicles, sterna and many other bones aren't
preserved either, so this is absence of evidence.>

  Not to mention that *Eoraptor* and *Herrerasaurus* are also preserved _sans_ 
gastralia, as in *Lewisuchus* and *Marasuchus.* So
this, also, doesn't say anything

<Makes at least 6 and at most 16 characters, versus 0 to 2 that argue 
otherwise.>

  I personally feel the number of ornithischian features are _far_ less :).

<I need not invite comments :-)>

  No, no need to invite comments. Other features that are primitive that I 
chose to disregard until discussion was entered after my
comments last week. The cervical vertebrae, like that of *Marasuchus* and 
*Lewisuchus,* bear large, craniocaudally elongated neural
spines, unlike the condition in "prosauropods," *Eoraptor,* or other basal 
theropods. The distal pubes are divided, unlike in
ornithischians, with a long and broad pubic apron, but this may be 
plesiomorphic as pelves of *Lewisuchus* and *Marasuchus* remain
very short in the pubis without the distal separation; otherwise, the pubes 
resemble "prosauropods" and, to a resemblance,
*Staurikosaurus,* which was almost the same age. An obturator foramen occurs, 
not a fenestra, thus apparently more primitive than
the condition in *Lesothosaurus,* *Eoraptor,* *Plateosaurus,* *Guaibasaurus* 
(inferred from broadly open margin of the pubis, as the
obturator plate is not preserved) etc. In *Agnostiphys, the ilium resembles 
that of *Thecodontosaurus,* whereas that of
*Silesaurus,* given the saddle-shaped dorsal margin, resembles only 
*Marasuchus,* *Lagosuchus,* and *Chindesaurus,* and does not
bear exceptionally elongated preacetabular (*Lesothosaurus*) _or_ 
postacetabular alae (*Thecodontosaurus,* *Lesothosaurus,*
*Agnostiphys,* *Guaibasaurus*), though the postacetabular ala is relatively 
longer than the preacetabular ala as in
*Thecodontosaurus,* but no other dinosaur bears the large triangular process of 
the preacetabular ala, or a dorsolateral margin
inclined more horizontally than vertically (~35 degrees above horizontal) nor 
does the dorsal margin contact the vertebrae above the
sacral ribs (they touch the transverse processes of the second and third 
sacrals) except among _derived_ ornithischians or theropods
(ankylosaurs, tyrannosaurs, birds, and ornithomimosaurs). The large lateral 
process of the postacetabular ala is absent in basal
dinosaurs except in *Herrerasaurus,* *Caseosaurus,* and *Chindesaurus.* The 
ankle is very primitive, but otherwise the condition of
the distal tibia and astragalus occurs in *Agnostiphys,* *Herrerasaurus,* 
*Eoraptor,* where the ascending process is very short,
triangular, pyramidal and not craniocaudally compressed into a plate, and 
almost centrally placed in fore-aft dimension. It is
further medial in placement compared to *Agnostiphys,* *Marasuchus,* 
*Pseudolagosuchus,* and *Herrerasaurus.* Finally, the  lesser
trochanter (lateral process on the craniolateral surface of the femur) is 
horizontal with a medially-positioned, vertical anterior
trochanter (anterior process on the cranial surface of the femur) and no 
expression of the greater trochanter and the femoral caput
except for a laminar ridge running proximally from the lesser trochanter; this 
condition is otherwise unknown in dinosaurs, or even
near-dinosaurs such as *Lagerpeton* or *Marasuchus,* and a fourth trochanter 
that is as proximal as the anterior trochanter.

   These features are common in more primitive, quadrupedal archosaurs, as well 
as the sigmoidal femur in cranial view. An analysis
of muscle position in reflection of its locomotory abilities, given the 
cursorial pes, would be an excellent piece of information
given the odd femoral anatomy. The conclusion should favor that this form is 
not a dinosaur based on the use of Ornithischia +
Saurischia. Dzik's usage of herbivorous adaptations to propose an ornithischian 
identity as being possible is erroneous on the basis
of comparison to other herbivorous amniotes (part of my own research), in which 
these features are convergent and the jaw anatomy
does not actually support any herbivorous adaptations to begin with.

<Do *Guaibasaurus* and *Saturnalia* really have completely closed acetabula 
like *Silesaurus*? :-o>

  Yes in *Saturnalia,* don't know in *Guaibasaurus*, as the ventral margin of 
the acetabular plate in the ilium of the latter is
incomplete, along with the cranial half of the ilium itself, and the acetabular 
portions of the ischium and pubis. Incidentally, by
contrast, *Agnostiphys* has a very arched aperture in the iliac acetabular 
plate similar to "prosauropods," which is remarkably
coincident with some of the lanceolate teeth recovered with the postcrania.

<Oh, and since when is *Chindesaurus* a poposaurid, as stated on p. 571? Fraser 
et al. say it has an ascending process on the
astragalus.>

  The references cited by Dzik discussing *Chindesaurus* all refer to it as a 
dinosaur. So this was curious to me, too. Rauhut (PhD
thesis) viewed the material as not diagnostic enough to be attributable either 
to dinosaurs, or with enough materiual to distinguish
it from some rauisuchians, like *Postosuchus.* It has a fully-closed 
acetabulum. The ilium of *Chindesaurus* that is so compared to
*Silesaurus* (as in Fraser et al., 2001, *Agnostiphys*) actually belongs to 
*Caseosaurus,* and the holotype of *Chindesaurus* only
bears a very partial ilium including a diagnostic postacetabular fragment. If 
there is anyone with cited research on *Chindesaurus*
as a poposaurid or other rauisuchian/basal crocodylomorphan, I (and David, I am 
sure)would love to see it.

  Cheers,

  Jaime A. Headden

  Little steps are often the hardest to take. We are too used to making leaps 
in the face of adversity, that a simple skip is so
hard to do.  We should all learn to walk soft, walk small, see the world around 
us rather than zoom by it.

  "Innocent, unbiased observation is a myth." --- P.B. Medawar (1969)