[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Horner and Goodwin on Triceratops

  In Horner and Goodwin's latest paper*, they offer a table of ten osteologic
features they suggest may lead to ontogenetic identity in *Triceratops* skulls.
They listed 31 specimens, of which ten could not be given a positive on one of
those ten features. These specimens range from "baby" to "subadult," but do not
include "adult."

  In the paper, Horner and Goodwin note that "Forster [_JVP_ 16:259-270]
re-evaluated *Triceratops* and recognized two species: *Triceratops horridus*
and *Triceratops prorsus*" and further mention that they "would expect the
cranial ontogenetic characters identified in our study for *Triceratops* to be
consistent, regardless of the number of species or genera accepted, because
cranial ontogeny is conserved in closely related taxa." This may have some
substantial bearing on the nature of the identifications, since the assumption
of similar ontogeny is presumed _a priori_ and was not tested specifically.
Indeed, since its first mention, the names of species do not come up again for
*Triceratops.* This is worth the testing, to some degree, to see if, even in
closely related taxa, suture closure, fusion of extraneous elements, and the
appearance and shape of sexually dimorphic or visual display structures would
differ from species to species, based on either 1) age of the species relative
to the other, 2) environment of the species relative to the other, or 3)
different maturation periods and phase intervals. These, as in (say) African
antelopes of even the same "genus," can vary to a diagnostic degree. I began by
evaluating the data as I would a data matrix, so this review is very limited in
scope and certainly not a comprehensive review in the least.

  First, the ten "characters" noted by Horner and Goodwin are the following:

Scalloped margin of the parietal and squamosal
Frill margin scalloped to wavy
Epoccipitals merged to frill
fan-shaped frill
Horns curved only posteriorly
Base of horns excavated internally
Inter-nasal fusion
Fusion of the nasal horn onto the nasals
Supraoccipital separated by the exoccipitals
Horns curved forward at the base

  If this were ran as a cladistic series using the skulls as operating
taxonomic units, these characters might thusly be coded:

1. Parietal/Squamosal, margin, shape: scalloping pronounced (0), slight
scalloping to wavy margins (1).
2. Epoccipitals, relative to the frill margin: separate (0), fused (1).
3. Frill, shape: low and semicircular in profile (0), tall and "fan-shaped",
more than 1/3 circular in profile (1).
4. Postorbital, horn, curvature: curved only caudally (0), curved anteriorly at
the base, then dorsally distally (1).
5. Postorbital, horn, internal cone-shaped basal excavation: absent (0),
present (1).
6. Nasals, fused: absent (0), present (1).
7. Nasal, horn: separate from nasals (0), fused onto nasals (1).
8. Exoccipitals, position: meeting at the midline above the foramen magnum (0);
separated by the supraoccipital (1).

  And if these characters were made into a matrix, the 31 specimens thus coded
would be rendered thus:

ucmp_154452     00000001
mor_652         00000000
mor_1199        10000000
ucmp_136306     10000000
mor_1110        10001000
ucmp_150234     10000000
mor_539         00000000
ucmp_137263     10010000
mor_1120        11111111
mor_699         10110001
ucmp_136092     101?0001
ucmp_137266     00110000
ucmp_173739     00010000
mor_1604        01010110
mor_004         01110110
mor_1625        01010110
ucmp_113697     01011100
ucmp_174838     01010110
ucmp_136589     01010000
ucmp_140416     01010000
ucmp_129205     01010000

  Included were only the specimens for which a coding was given in Horner and
Goodwin's table, but this still left MOR 652, considered a "baby" by the
authors, without any positive coding. As such, it would be the "outgroup" by
default, though the resultant trees all treat the two included "babies" in the
table as closest to the root anyway, along with a variety of other specimens,
due to the singularity of their given codings.

  This matrix, thus produced, resulted in 25 different trees, a testament to
this matrix's completeness and thorough handling of the characters. The
consistency index (CI) comes out at a flat 0.5, as does the homoplasy index.
These are not encouraging numbers, but the matrix is so much smaller than the
sampled taxa! I ran the tree under DELTRAN optimization and with all characters
unordered, and recieved a strict consensus tree where only two real patterns
emerged: MOR 699 and UCMP 136092 always grouped together (both of these listed
as "subadult"), and all specimens listed by Horner and Goodwin as "adult"
grouped together as a unit with MOR 1120 (the winner of the matrix as the only
taxon to score on all characters with a "1", listed as a "subadult") paired
consistently with MOR 004. The latter was the "quintessential" poster child for
the adult phase in Horner and Goodwin's view, and this might be saying

  As I look at the resultant trees, I find them unsatisfying for the purpose,
but note that the ontogeny is not measured against skull length, provenance, or
age of the strata in which they were recovered. This information, when checked,
might produce a new set of constraints that may actually allow differentiation
of species, or destroy the species limits, or find that as species, these taxa
do not age any differently. That the "iconic" adult, MOR 004 has very short,
nearly unrecurved horns and a prominent nasal horn, suggests it might be a VERY
old adult, which has had some time to grow its schnoz, and the postorbital horn
cores began to dissolve with onset of osteoporosis (I'm throwing this out from
my butt, so beware as this is based only very partly on Horner and Goodwin's

* Horner, J. R. & M. B. Goodwin. 2006. Major cranial changes during
*Triceratops* ontogeny. _Proceedings of the Royal Society, B_ 273:2757-2761.

  "This is the first cranial ontogenetic assessment of *Triceratops*, the
   well-known Late Cretaceous dinosaur distinguished by three horns and a
   massive parietal?squamosal frill. Our analysis is based on a growth series
   10 skulls, ranging from a 38 cm long baby skull to about 2 m long adult
   skulls. Four growth stages correspond to a suite of ontogenetic characters
   expressed in the postorbital horns, frill, nasal, epinasal horn and
   epoccipitals. Postorbital horns are straight stubs in early ontogeny, curve
   posteriorly in juveniles, straighten in subadults and recurve anteriorly in
   adults. The posterior margin of the baby frill is deeply scalloped. In early
   juveniles, the frill margin becomes ornamented by 17?19 delta-shaped
   epoccipitals. Epoccipitals are dorsoventrally compressed in subadults,
   strongly compressed and elongated in adults and ultimately merge onto the
   posterior frill margin in older adults. Ontogenetic trends within and
   growth stages include: posterior frill margin transitions from scalloped to
   wavy and smooth; progressive exclusion of the supraoccipital from the
   magnum; internal hollowing at the base of the postorbital horns; closure of
   the midline nasal suture; fusion of the epinasal onto the nasals; and
   epinasal expansion into a morphologically variable nasal horn. We
   that the changes in horn orientation and epoccipital shape function to allow
   visual identity of juveniles, and signal their attainment of sexual

  This paper is a follow-up to Goodwin et al. 2006, _JVP_ 26(1):103-112,
published earlier this year, describing the skull of UCMP 154452.

Jaime A. Headden

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

Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around