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

New paper on Caudipteryx & cursoriality in Nature

Just published:

Jones, T.D., J.O. Farlow, J.A. Ruben, D.M. Henderson & W.J. Hillenius.
2000.  Cursoriality in bipedal archosaurs.  Nature 406: 716-718.

Okay, so, its an interesting paper.

The two main analyses conducted:
Morphometric plots of hindlimb length versus trunk length in bipedal
ornithopods, non-avian theropods (to the exclusion of _Caudipteryx_),
_Caudipteryx_, and modern cursorial ground birds.  There are two such plots:
one where the total hindlimb (THL = femur, tibia/tibiotarsus, and
metatarsalIII/tarsometarsus) of all taxa are plotted, and a second in which
"effective hindlimb" length (EHL = a concept from the forthcoming Farlow et
al. paper in American Zoologist) (same measurements for non-avian taxa, but
for birds and for _Caudipteryx_ only the epi- and metapodium are used).

In this plot, _Caudi._ and birds plot higher (longer legged) than typical
striding dinos for the THL plot, and all the EHLs plot along the same line.

Second analysis is a center-of-mass study.  The center of mass in
Caudipteryx was found to be  2.3 times further forward of the acetabulum
than in Deinonychus.

Their conclusions: these data, coupled with the lack of a fourth trochanter
in _Caudipteryx_, suggests to them that this critter had a bird-like rather
than a plesiomorphic style of gait.  They suggest three alternative
hypotheses to explain this: 1) that _Caudi._ was an "unusual theropod
dinosaur" that abandoned dino-style running and independantly derived an
avian-style running; 2) that _Caudi._ is a secondarily flightless non-avian
dino; 3) that _Caudi._ is a secondarily flightless bird (i.e., derived from
birds after the origin of _Archaeopteryx_).  In fairness, they do not insist
on the last possibility, but do say that it is interesting that the only
unquestionable feathers among supposed non-avian dinos also are found on a
critter with the same locomotory style as known secondarily flightless

Okay, fair enough.

However, there are a couple of major methodological flaws.  I will address
them here, for the sake of sharing this information now, but rest assured I
will also be addressing it more appropriate peer-reviewed fora.

Problem I (the BIGGEST problem): The use of restorations and reconstructions
rather than fossils for their data.  In seven out of their 18 theropods,
either the hindlimb is incomplete and missing one or more segment (for
example, they given measurements for _Carnotaurus_ tibiae and metatarsi, but
only the proximalmost part of the tibia of this dinosaur is known, and at
present NO abelisaurid metatarsi are described!!) or the dorsal column is
only known by less than half the vertebrae (in at least one specimen they
use, I cannot confirm that a single dorsal vertebra was recovered: they are
thus plotting an X-Y value where X is totally unknown!).

In their Methods section they state that they only used specimens that they
were confident in their reconstructions.  Sure, sounds good, but they are
apparently missing the circularity of argument here.  Their goal is to
discover the underlying scaling pattern between hindlimb and trunk length.
In order to do this, they use imaginary measurements which are themselves
scaled by the artist from other known specimens.  To no real surprise, these
pre-scaled data fall along the same scale as other specimens.

Would exclusion of these incomplete specimens affect their scaling pattern?
I don't know yet, but I am going to do the scientific thing and test it.

Problem I, subsection B: Related to this, the _Deinonychus_ they use is Greg
Paul's composite of YPM 5232 and AMNH 3015 (on p. 363 of PDW).  Leaving
aside the fact that this is an out-of-date restoration (forelimbs would be
differently done these days, for instance), this is not a specimen: it is a
composite of two individuals.  Furthermore, even put together the two
specimens are nowhere near as complete as restored.

Ironically, just lower on the same page is a restoration of GI 100/25: the
Fighting _Velociraptor_.  This would have been a much better choice all
around: it is a single, complete specimen; by their own criteria (ossified
uncinate processes, ossified sternum, etc.) it is an adult; and it is much
closer in size to _Caudipteryx_ than is _Deinonychus_, and thus reduces the
possible effects of allometry on center-of-mass (and instead concentrates on
differences in hindlimb anatomy).

Problem II: Exclusion of juveniles.  The authors chose not to plot specimens
they consider juveniles (although they include _Eustreptospondylus_, known
at present only from an immature specimen), on the basis that previous
studies show that juvenile specimens of theropods have proportionately
longer hindlimbs than do adults.  True, but those same studies (including
some by Gatesy and by myself) find that juvie specimens of larger non-avian
theropods tend to plot about the same spot as adult specimens of small
theropods of the same body size.  Thus, rejection of the use of juveniles
may not be justified (let's test it and find out), and would open up the use
of additional specimens to replace those that must be removed because they
are wanting in major hindlimb elements or large sections of the dorsal

Additionally, the authors do point out in the Methods section that the
troodontid _Sinornithoides_ and the dromaeosaurid _Bambiraptor_ (hate that
name...) plot among the birds and _Caudipteryx_ rather than with the typical
non-avians.  However, they do not show these in the main plot, as they are
juvenile specimens rather than adults.  As mentioned above, though, we do
not at present know that typical non-avian theropods *would* plot off the
main line (and previous studies suggesting that they would indeed plot on
the main line).

Further information: What to do with the results.  In fairness, the
phylogenetic position of _Caudipteryx_ has not been as adequately tested in
currently published studies as I'd like to see.  The initial Ji et al. study
included only one non-avian group (Velociraptorinae), and that only as an
outgroup, so that a position of _Caudipteryx_ with any other clade of
non-avian dinosaur was not a possible result.  In Sereno's 1999 analysis in
paper, the Tetanurae study (see the supplementary data to that paper)
includes plently of non-avian groups, but Aves is included only as a single
OTU, so a result where _Caudipteryx_ was inside the _Archaeopteryx_-bird
clade was not possible.

There are forthcoming papers in press in which an avian or a non-avian
position of _Caudipteryx_ are both possible.  However, these do not
incorporate the new hindlimb proportion data Jones et al. present.

So, the authors of this study suggest that the possibility of a secondarily
flightless non-avian _Caudipteryx_ deserves closer scrutiny.  Too bad we
don't have a methodology for testing alternative hypotheses of phylogenetic
position... Oh, wait... yeah.


                Thomas R. Holtz, Jr.
                Vertebrate Paleontologist
Department of Geology           Director, Earth, Life & Time Program
University of Maryland          College Park Scholars
                College Park, MD  20742
Phone:  301-405-4084    Email:  tholtz@geol.umd.edu
Fax (Geol):  301-314-9661       Fax (CPS-ELT): 301-405-0796