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

Czerkas' book- Analysis and Criticisms



Yesterday I got Feathered Dinosaurs and the Origin of Flight, and I must say
that though the specimens are excellent, the writing is perhaps the worst
I've read regarding the dinosaur-bird "debate".  At least the ABSRDists are
consistant (if wrong), and have an understandable (though flawed)
methodology.  I'll proceed in the order the papers are in-

Czerkas and Mickelson, 2002. The first occurence of skeletal pterosaur in
Utah.
Did you know "The fossil record of flying reptiles known as pterosaurs are
exceedingly rare in the western hemisphere."?  That's the opening sentence
to this chapter.  Grammatical errors aside, where do they think the Santana
Formation is located?  The Eastern hemisphere?  Czerkas and Mickelson also
give the impression that people have argued pterosaurs were rarer than usual
in America (or the Morrison Formation in particular), though as far as I
know paleontologists have always (correctly) believed this to be due to the
Morrison's bias against preserving small vertebrates.  Still the authors use
much of the introduction and discussion explaining how Utahdactylus is
important in demonstrating pterosaurs were not absent from the Northen
hemisphere and such.  They have an obsolete dichotomous view of pterosaurs
being divided into Rhamphorhynchoidea and Pterodactyloidea, as opposed to
modern work by Unwin, Peters and others that shows the first was a
paraphyletic grade.  Utahdactylus itself is based on cross sections and
natural casts of bone fragments, which I have a hard time believing are
diagnosable.  The "diagnosis" consists of it having some rhamphorhynchoid
plesiomorphies (elongate tail with zygopophyseal rods, short cervical
vertebrae) and supposedly lacking the longitudinal groove on manual digit IV
phalanges that is typical of rhamphorhynchoids.  The latter is based on
"cross sections and portions of shafts that may represent the radius and
ulna from both limbs, as well as the phalanges from the fourth digit.
Unfortunately, all of the assignable elements are so incomplete as to
prevent positive identifications ...".  While it may have actually lacked a
groove, this seems far from certain.  If it did, the first idea I have is
that it could be a derived "rhamphorhynchoid" that was closer to
pterodactyloids in this character.  Czerkas and Mickelson seem not to have
thought of this, and instead suggest it "would appear to reflect a more
generic condition than ordinarily found in rhamphorhynchoids and
pterodactyloids."  The actual description, though more than a page long, is
almost completely lacking in morphological information.  I'd like to see
what Unwin, Bennett, Peters and such would think of this specimen and the
next.

Czerkas and Ji, 2002. A new rhamporhynchoid with a headcrest and complex
integumentary structures.
Here the old rhamphorhynchoid-pterodactyloid dichotomy shows up again.  This
time, Pterorhynchus is special for having "the first example (of a cranial
crest) known to exist among the long-tailed pterosaurs", despite the fact
Austriadactylus (Dalla Vecchia et al., 2002) was described earlier, and is
even mentioned as having a crest in the discussion.  The specimen is well
preserved, but some parts (eg. posterior skull) are unprepared.  Again, the
description is not very useful at all when it comes to skeletal anatomy,
though the soft tissues are well described.  Unfortunately, the photos are
not helpful in discerning skeletal anatomy either.  And though it's assigned
to the Rhamphorhynchidae, there is no reasoning for this and virtually no
morphological comparison to other taxa.  There are some interesting soft
tissues associated with this specimen.  The cranial crest has a small bony
base, but is mostly keratinous.  It's covered with tiny longitudinally
pointed ovoid scales, but has layers of longitudinal ridges basally.  There
are also short fibers at the base of the crest, and vertical stripes that
seem to represent the original color pattern.  The body covering is
filamentous as well, arranged in V-shapes that the authors believe indicates
plumulaceous feathers were present.  There are lots of tiny dots that
flouresce in ultraviolet that they interpret as follicles, along with thick
hollow circles that are seen as calami.  These views seem reasonable enough,
certainly moreso than Longisquama's "parafeathers".  The tail has a long and
slender vane, which is also covered with filaments.  A more questionable
conclusion is that there are tiny plumulaceous feathers on the wing
membrane, perpendicular to the actinofibrils (with occasional calami
possibly preserved).  It's possible, but though I might be able to see
fibers, I can't see the supposed organization.  None of the feathers or
fibers are said to have barbules.  The discussion reveals that the authors
believe the feathers of Sinosauropteryx are homologous with those of
pterosaurs and birds, and that they originally evolved for insulation.  This
is rather logical, but Czerkas and Ji's statement that birds and pterosaurs
must have shared an arboreal ancestry is not.  Why must feathers have
evolved in the trees?  Why not in terrestrial creatures first?  The authors
feel the primary function of cranial crests was aerodynamic, then camoflage,
and finally display.  I don't know about this.  What about species like
anuroganthids, with such short snouts that probably didn't have crests?  If
the crest was vital for flight, how did the uncrested taxa manage it?

Czerkas and Xu, 2002. A new toothed bird from China.
The authors claim that "the name Archaeoraptor has become a nomen nudum",
which I was unaware of.  They renamed it because the "terminology is not
appropriate" and "to more accurately reflect the animal's ability".  From
what I understand, these are not valid reasons to rename a taxon.  Czerkas
and Xu use the outdated stem-based definition for Ornithurae, which roughly
corresponds to Chiappe's (2001) Ornithuromorpha and Sereno's (1999)
Euornithes.  The diagnosis is flawed, and I believe Archaeovolans is a
junior synonym of Yanornis (though not Songlingornis).  This description is
actually the best in the whole volume.  The problem is that Czerkas and Xu
seem to have forgotten birds intermediate between Archaeopteryx and
ornithurines are known, except for their few references to Confuciusornis.
Whenever a character more primitive than ornithurines is found, it is
described as being similar to Archaeopteryx.  One would think that somewhere
in the paper would be comparisons or references to enantiornithines,
Songlingornis, Patagopteryx or even the supposed "earliest ornithurines" of
ABSRDists- Chaoyangia and Liaoningornis.  But no, comparisons to related
taxa were apparently not deemed important.
My preliminary description-
http://www.cmnh.org/dinoarch/2001Apr/msg00444.html is largely correct,
though a few additions can be made.  The premaxilla has four teeth.  The
dentary has at least eighteen teeth in distinct sockets.  The teeth lack
serrations and have constricted roots.  A slender bone said to be a palatine
can be seen in x-rays.  If correctly identified, it is triangular.  The rest
of the skull is much too crushed to interpret unambiguously.  Cervical ribs
are said to be "not completely ossified" to the vertebrae when viewed in
ultraviolet.  A dorsal vertebrae is said to be opisthocoelous.  At least ten
ribs are preserved, and sternal ribs are present as well.  Ossified
uncinates could not be identified, but we know the uncertainty involved with
judging their presence in imperfect specimens.  The element I identified as
either a scapula or metacarpal II is metacarpal II.  The coracoid has a
procoracoid process and concave facet for the scapula.  Czerkas and Xu make
a big deal out of the facies articularis ulnaris (ridge on proximolateral
metacarpus), which is absent in Yanornis.  They state it is absent in
Archaeopteryx (as it's primitive), but don't realize it's also absent in
Confuciusornis, Protopteryx, Longipteryx, enantiornithines and Patagopteryx.
Contra my earlier description, the semilunate is fused to the second
metacarpal.  The metacarpus is unfused.  An intermedium(?) is preserved
proximal to the third metacarpal.  The illustration I used previously was
wrong in portraying the first digit as extending past metacarpal II,
eliminating the only character I noted previously that was too primitive for
an ornithuromorph.  Phalanx II-2 is preserved, and shorter than II-1, while
only a single phalanx was present on digit III.  There are contour feathers,
and shafts of primaries.  Unfortunately, the questionably associated
hindlimb elements are not illustrated, or described in detail.  The
metatarsus is fused only proximally though, which would make them from a
non-ornithuromorph.  It does not resemble that of Yanornis.  It is thus
incorrectly associated, though what it belongs to awaits better illustration
or description.
Czerkas and Xu state that the upper teeth being restricted anteriorly is
unique among Early Cretaceous birds, but they couldn't even identify the
maxilla, so this statement is baseless.  A lack of phylogenetic
understanding is illustrated by the last sentence- "The mixed developmental
status of characters associated with flight may suggest that Archaeovolans
was possibly eqipped with a pygostyle, but it remains unknown as to how long
or short the tail actually was and whether or not it possessed any
transitional traits towards becoming like that of modern birds."  Even with
my data limited to a skeletal outline, I narrowed down this specimen's
phylogenetic position to well within pygostylous birds.  And that was over a
year ago.  Czerkas and Xu believe this taxon to be an ornithuromorph, you'd
think they would realize it near certainly possessed a pygostyle.  Overall,
I think the authors were trying to make this specimen seem more primitive
than it really is.  Coincidentally, I think the model of this taxon is
fairly inaccurate, with the beak too short, and the hands too generic and
primitive.  Still, the paper was not as bad as the next.

Czerkas and Yuan, 2002. An arboreal maniraptoran from Northeast China.
Here the real fun begins.
First, the authors seem to have gotten it into their heads that this strict
dichotomy exists among paleontologists.  With ABSRDists being completely
against a terrestrial origin of flight, and BADDists being completely
against an arboreal origin of flight.  Of course, the latter isn't true,
especially in recent years.  Just look at Xu et al. (2000) or any number of
talks at last years SVP.  What's even more confusing is that the authors
reference Paul and Olshevsky numerous times, but never use them as examples
of arboreal origin BADDists.  They only reference them when it suits their
interests, such as saying that neoflightless deinonychosaurs were previously
hypothesized to exist.  Oh yes, apparently BADDists believe that
Archaeopteryx was "primarily a terrestrial animal with little or no
capability of flight".  One of the papers cited as supporting this view is
Burgers and Chiappe (1999), which was specifically about proving
Archaeopteryx could gain enough speed while running to fly.  Another logic
problem Czerkas and Yuan have is their inability to concieve of an arboreal
theropod.  Apparently they find it utterly impossible for theropods to
climb, which is difficult to understand.  How can they believe evolution to
be limited in such a way?  Yet such unscientific statements as the one below
are strewn throughout their paper- "The specimen described below challenges
the popular and widely held dogma of a dinosaur/bird relationship by
presenting the strongest evidence discovered so far that the ancestors of
birds were uniquely adapted to an arboreal lifestyle...".  Apparently
theropods "by definition" do not climb.  Since when are clades defined by
behavioral limits?  I don't see how Czerkas and Yuan can get away with this
nonsense.  Theropods are seemingly also defined by the second manual digit
being longest, but at least Czerkas seems able to comprehend of a reversal
occuring here.  Yet another confusing aspect is that though they claim to be
against a dinosaur ancestry for birds, they place birds in the Saurischia.
What are saurischians if not dinosaurs?  Czerkas and Yuan state "Whether
these saurischians can be considered dinosaurs depends on the definition of
what a "dinosaur" is."  What the hell is that supposed to mean?  When have
saurischians ever not been considered dinosaurs?  And they continue, "...
dinosaurs may be related to the avian lineage in being derived from a common
ancestor within the earliest stages of saurischian archosaurs."  If that was
true, birds would be dinosaurs.  Either Czerkas and Yuan are deliberately
trying to avoid BADD, by defining "dinosaur" as some polyphyletic group that
excludes birds, or they don't know what they're talking about.  And despite
the authors' unquestioning assignment of Scansoriopteryx to the Saurischia,
there is no explanation why they put it in that clade, except that the
pelvis is non-opisthopubic (which is a primitive trait).  Czerkas and Yuan
state there are "massive reversals secondarily resembling primitive
characteristics" that would have to take place if Scansoriopteryx were a
theropod.  Scanning the text for these supposed reversals reveals very few.
The crescentric pterygoid-quadrate contact is different from most theropods,
but Archaeopteryx and basal saurischians like Herrerasaurus resemble
theropods in this area.  So a reversal had to happen sometime regardless of
whether Scansoriopteryx is theropod or a basal saurischian ancestral to
birds.  The sacrals are said to be wider than theropods, but Microraptor is
comparable.  Of course, they consider deinonychosaurs birds, but one could
note Velociraptor has narrow sacrals more typical of theropods.  So again,
reversals had to happen somewhere, regardless of phylogeny.  The narrow
pubic peduncle is said to be a primitive character, but theropods like
Avimimus, alvarezsaurids and therizinosaurids developed narrow pubic
peduncles, as did birds like Patagopteryx.  So if Patagopteryx could reverse
the state, why not Scansoriopteryx?  The partially enclosed acetabulum would
likewise have to change states sometime, as some deinonychosaurs have it
(eg. Unenlagia), and others do not (eg. Deinonychus).  The supposedly low
puboischial ratio could be caused by the pubes being incomplete, as the
proximal shafts are unpreserved.  Even if true, it has to reverse before
Archaeopteryx and deinonychosaurs come along.  I don't believe the supposed
clavicle is complete, and find it likely to be part of a furcula.  The third
manual digit is considered extremely important by Czerkas, despite the fact
phylogenetic signal is less likely to be retained in a hypertrophied organ
specialized for a certain behavior.  Supposedly primitive characters are the
long digit III and metacarpal, and the fact III-3 is the shortest phalanx in
that digit.  The former is present in some ornithomimids, while the latter
is found in Ingenia.  These examples show the characters can evolve from a
normal theropod state, as more basal ornithomimosaurs and oviraptorosaurs
have the usual theropod condition.  The supposedly robust fibula is just
part of the crushed tibia in my opinion.  So the "numerous primitive
characteristics throughout the body" just don't exist.  Paradoxically, the
best piece of evidence that Scansoriopteryx is not a theropod, the ischial
structure, is not mentioned in that context anywhere, but instead only in
the context of being more primitive than Archaeopteryx.  It seems Czerkas
and Yuan have no problem with characters chaging from some states to others,
but consider the reverse impossible.  It's like some odd version of Dollo's
Law.  And sometimes the authors refer to Scansoriopteryx as a theropod- "...
which is parallel in development to all theropods except Scansoriopteryx."
or "The evidence for such an unequivocal arboreal theropod...".
Protarchaeopteryx and Caudipteryx are seen as birds.  I'm guessing this is
due to Czerkas' crazy definition of "bird", which will be discussed below.
Both BCF and neoflightlessness are supported, but the authors show a lack of
understanding when they claim neoflightlessness is contrary to recent
phylogenetic analyses.  Though phylogenetic analyses can support
neoflightlessess, they can not disprove it because potentially volant taxa
that would plot more basally could be unknown at the time.  Yet another
problem is that Czerkas and Yuan think Scansoriopteryx completely disproves
the cursorial origin of flight, because it is arboreal.  Did it ever occur
to them that Scansoriopteryx could be neoflightless, taking up an arboreal
niche after its ancestors gained flight cursorially.  Now, I prefer the
arboreal or Dial hypotheses myself, but the situation I outlined isn't
implausible.  The authors also believe Scansoriopteryx is ancestral to
maniraptorans (though they call it a maniraptoran in the next chapter),
which they say proves maniraptorans became cursorial in parallel to
theropods.  First, note Czerkas and Yuan provide no shared derived
characters with non-avian maniraptorans, so why they place it there is
unknown.  After reading the Cryptovolans paper, their reasoning becomes
clear (as explained below).  The authors don't understand the concept of
parsimony as it relates to reversals, as they say the long third manual
digit of Scansoriopteryx proves theropods and birds (including
maniraptorans) reduced that digit convergently.  Assuming theropods and
birds were sister groups (as shown in their phylogenetic tree) the long
third digit in Scansoriopteryx could be a reversal of the "short third
digit" character uniting the groups, or primitive compared to convergently
developed short third digits in the two groups.  PAUP would find the
possibilities equally likely, ACCTRAN would find the first, DELTRAN the
latter.  So why aren't birds considered theropods if they are their
sistergroup anyway?  Because of Czerkas' definition of theropod, which
states they must not climb or have the third manual digit be longest.  For
all the negatives of Czerkas and Yuan's paper, we do all agree
Scansoriopteryx is closely related to, but more primitive than,
Archaeopteryx.  And I should mention the paper is wonderfully illustrated,
with close-ups and line drawings of everything.  The skeletal reconstruction
is much better than ABSRDist attempts at Caudipteryx and such, but still has
problems.  The tail is only twenty-two caudals long in the reconstruction,
despite the fact the twenty-second caudal shows no sign of being the last.
The infratemporal fenestra is shaped incorrectly, and ironically the ischium
is reconstructed improperly but looks more theropodan than it actually does.
Some morphological details can be added to what I described earlier (
http://www.cmnh.org/dinoarch/2002Sep/msg00668.html ).  The integument
consists of plumulaceous feathers, with round features interpreted as
calami.  The feathers are especially long on the ulna and metacarpus.  This
specimen is almost exactly the same size as the Epidendrosaurus one (various
elements .95-1.08 times the size of Epidendrosaurus, average 1.00; my
previous measurements were based on the scale bar).  The quadrate is single
headed.  The retroarticular process is not elongate, further evidence
against oviraptorosaurian affinity.  The posterior dorsals are amphiplatyan,
with no pleurocoels or lateral fossae.  This is similar to troodontids and
Microraptor, as well as some basal birds like Sapeornis.  The vertebrae have
unfused neurocentral sutures and the sacrals are unfused, showing the
specimen is not an adult.  The lack of a brevis fossa is consistant with
either a paravian or enigmosaurian affinity, but not something more basal
like ornithomimosaurs or Ornitholestes.  The acetabulum is partially closed,
as in bird-like eumaniraptorans.  In this feature, and the concave
posterodorsal ilial outline (associated with a tubercle in other
eumaniraptorans), Scansoriopteryx strongly resembles Unenlagia.  Twenty-two
caudals are preserved, though I estimate between thirty and thirty-five were
originally present.  Transverse processes only extend to the ninth caudal,
while neural spines extend to the eighth caudal.  This level of reduction is
only known in avialans and Microraptor.  So my phylogenetic hypothesis
remains the same as before, and is strengthened by new information.  I
believe Scansoriopteryx is a eumaniraptoran, best placed in the basal
Avialae.

Czerkas, Zhang, Li and Li, 2002. Flying dromaeosaurs.
It's here we see why Czerkas wants dromaeosaurs and other maniraptorans to
be birds.  He states "This distinction (being derived from flying ancestors)
further based on having feathers which enabled flight is an essential
prerequisite in the definition of a bird."  So basically, if something is
volant or neoflightless and has feathers, it MUST BE a bird to Czerkas.  And
hence, maniraptorans MUST BE birds.  Now, this wouldn't be too crazy if
Czerkas just had this as an apomorphy-based definition of Aves.  It would be
like Avialae sensu Gauthier and de Queiroz (2001).  The problem is that this
isn't an apomorphy-based clade, it's an apomorphy-based topology.  For if
any "theropods" are shown to be neoflightless, they are immediately shoved
out of whatever position they had in Theropoda, and placed as a bird derived
from "saurischian archosaurs".  Morphology alone cannot distinguish
theropods and birds in Czerkas' mind.  Indeed, he writes, "In addition to
the dromaeosaurs, the removal of other bird-like, or "non-avian dinosaurs"
from the Theropoda is also almost a certainty.  But prior to this happening,
the distinction needs to be made on whether the flightless forms have really
passed through an ancestral stage that included flight, or not.  This
distinction of having an ancestor that could truly fly is paramount in the
definition of a bird.  The problem of convergence and reversals may make
this determination difficult or impossible in many cases."  That's right,
impossible.  First of all, if theropods and birds can be so similar that
it's impossible to tell them apart, why are they kept separate?  Because
Czerkas, for some illogical reason, believes theropods could not have
possibly led to arboreal or volant forms.  I'm not sure what mechanism
Czerkas invokes that causes this evolutionary block, though it's certainly
unscientific.  I would reiterate that Czerkas would have to keep on calling
more and more primitive theropods birds, but he's created an easy way out of
this problem.  You see, he can ignore morphology.  Sure, ornithomimosaurs
LOOK a hell of a lot like basal maniraptorans, and share plenty of
characters with them, but because ornithomimosaurs weren't neoflightless,
that doesn't matter.  Czerkas KNOWS theropods cannot ever evolve flight, so
only maniraptorans will ever be birds.  I'm extremely disappointed such an
unscientific theory got in print.  At least the ABSRDists keep up a list of
features distinguishing dinosaurs from birds, even though the characters are
misapplied or fictional.  For someone to just say they could be identical,
but are not closely related, is just sad.
Czerkas et al. say that NGMC 91 (cf. Sinornithosaurus) was possibly
cogeneric with Cryptovolans, but this is wrong because it lacks the
distinctive manual character of Cryptovolans, where III-1 is longer than
III-3.  They also believe cf. Sinornithosaurus' primaries were just
unpreserved, not actually absent.  I don't see a reason this must be so,
especially in a subadult.  In this paper, as with Pterorhynchus, the
description is brief to a fault.  The skull is undescribed, though the teeth
are illustrated.  They have serrations only posteriorly, which is only known
in basal coelurosaurs and eumaniraptorans.  No presacral vertebrae are
described, but the tail has 28-30 vertebrae with prezygopophyseal and
chevron rods.  The latter are only known in dromaeosaurs, while the number
of caudals is intermediate between Deinonychus (~35-40), and Microraptor
(26) and cf. Sinornithosaurus (~20).  Uncinate processes are present.  The
furcula is nearly identical to Bambiraptor, but with a smaller
interclavicular angle.  The scapulacoracoid is similar to Sinornithosaurus,
differing in having a lower more extensive acromion and shorter posterior
process.  The sternum is similar to Sinornithosaurus', but fused together
with a central ridge.  I think I can see a stout lateral process
posteriorly.  Sternal ribs are present.  The arms are 83% of hindlimb
length, with subequal radii, ulnae and humeri.  The ulnofemoral ratio is
supposed to be virtually the same as Microraptor, which would be .66.  The
radius is thinner than the ulna, which is bowed.  The manus is almost
identical to cf. Sinornithosaurus except that phalanx III-3 is shorter than
III-1 (III-1/III-3 ratio 1.39).  Czerkas et al. make a big deal about this,
giving us yet another definition of Theropoda- those saurischians whose
fingers have "reduced proximal phalanges and elongate penultimate
phalanges".  Of course, Ingenia slipped by them somehow.  The authors say
this character of Cryptovolans is plesiomorphic, making it "closer to
Scansoriopteryx than any other dromaeosaur or maniraptoran."  This would
make dromaeosaurs paraphyletic to other birds of course, which Czerkas et
al. seem to ignore.  Interestingly, Sinornithosaurus has a very large
III-1/III-3 ratio (.81), and Microraptor an even larger one (.98).
Deinonychus and Velociraptor are more similar to other theropods in this
regard.  The pelvis is said to be like Sinornithosaurus- opisthopubic and
with a short ischium.  The tibiofemoral ratio is said to be 10% less than
Microraptor, which would be 1.18.  For reference, the tibia of the holotype
is about 131 mm, and that of the referred specimen LPM 0159 is about 122 mm.
Unfortunately, there are no measurements given.  There is an enlarged second
pedal ungual.  The remiges are asymmetrical, and the retrices are confined
to the distal tail.  I currently don't have an opinion on whether
Cryptovolans could fly or whether it had elongate tibial and femoral
feathers.  Cryptovolans looks to be a basal dromaeosaur most closely related
to Microraptor and Sinornithosaurus.

Czerkas and Ji, 2002. A preliminary report on an omnivorous bird from
Northeast China.
This paper is probably the best in the volume.  The authors find it
important that the skull is so much like Caudipteryx, which I agree is odd.
If all I had to go on was the skull x-ray, I would probably consider it a
caudipterid.  They also believe Eoenantiornis' (
http://www.cmnh.org/dinoarch/2000Dec/msg00357.html ) skull is similar, but
this taxon lacks the oviraptorosaur-like characters of Omnivoropteryx
(premaxilla and anterior maxilla with low ventral margin; decurved dentary;
enlarged external mandibular fenestra).  The short snout is probably a
juvenile trait, as judged by the sternal morphology.  I don't see any of the
"highly derived characters" that supposedly unite Caudipteryx,
Omnivoropteryx and Eoenantiornis.  Frankly, I'm confused as to exactly what
the authors are saying about the relationships of Caudipteryx, ornithomimids
and oviraptorids.  Most of the time, Caudipteryx is presented as a bird, but
they say "... it may not be possible to demonstrate conclusively whether
Caudipteryx was a flightless precursor of birds, or if it was a secondarily
flightless descendent of them."  According to Czerkas' beliefs, Caudipteryx
wouldn't be a theropod in either of those situations.  A couple paragraphs
later, Caudipteryx is referred to as a theropod dinosaur- "Regardless of
their bird-like characteristics, these theropod dinosaurs [including
ornithomimids and oviraptorids] are otherwise too specialized to have been
directly ancestral to any volant bird, whereas their possession of avian
characteristics is not so problematic if they were inherited from a more
volant avian ancestor."  Here, Czerkas contradicts his theory in the last
chapter, which maintained that by definition, theropods could not be
ancestral to volant birds.  Also note by including ornithomimids, Czerkas
might be placing all maniraptoriformes inside Aves, not just all
maniraptorans.
My earlier analysis of Omnivoropteryx (
http://www.cmnh.org/dinoarch/2002Sep/msg00668.html ) can be supplemented
with some additional information.  The premaxillae are unfused, placing it
below Pygostylia.  There are four procumbant premaxillary teeth with
constricted roots, like many basal birds.  The maxillae and dentaries seem
toothless though.  There is probably one phalanx on manual digit III.  I
still believe it is best placed as a non-pygostylian avialan.

I recommend you all buy the book.  Not for the discussion, but rather for
the illustrations of the wonderful specimens it contains.

In conclusion, I think this book is a fine example of one of Czerkas and
Yuan's quotes- "... the complexities (the incomplete fossil record) creates,
especially when objectivity is obfuscated by the misleading confines of a
preconceived dogma."

Mickey Mortimer