Kurochkin 2006 critique

[Michael Mortimer <mickey_mortimer111@msn.com>]

Kurochkin, 2006. Parallel Evolution of Theropod Dinosaurs and Birds. 
Entomological Review, Vol. 86, Suppl. 1, pp. S45?S58.

Kurochkin's latest (2006) paper is an interesting entry into the BAND camp.  
In a way, it is the antithesis of MANIAC (Maniraptorans Are Not In Actuality 
Coelurosaurs) as endorsed by Feduccia, Martin and Czerkas.  Instead, 
Kurochkin believes Archaeopteryx and enantiornithines are coelurosaurian 
dinosaurs, while confuciusornithids, Protoavis and euornithines are 
descended from a more basal archosauromorph.
As could be expected, Kurochkin rejects cladistics.  He cites Dodson (2000), 
with such wonderful rationales as "parsimony (economy), contradicts the 
essence of the evolutionary process in nature, because evolution is wasteful 
in every respect."  Er, if this were true, morphology wouldn't reflect 
phylogeny at all.  Also presented are falsehoods like "Cladistists exclude 
stratigraphy, embryology, physiology, ecology, and biogeography from 
consideration and take into account only formalized morphological 
characters, irrespective of space and time;."
On the plus side, Kurochkin recognizes theropods have numerous birdlike 
characters and correctly understands the BAD viewpoint for the most part.  
He lists some of the problems BADists have with BANDists, but waves them 
away by referencing his own 2001 paper and Feduccia's (1999) fossil bird 
book.  The later certainly doesn't address the problems satisfactorily, 
though I haven't read Kurochkin (2001), so can't comment further on this 
matter.  He does seem to think the contradictory conclusions of various BAD 
cladograms (e.g. the varying sister taxa of birds) somehow weaken the BAD 
hypothesis, which doesn't follow from logic.  Also, I'm not sure where 
Kurochkin has been in the last few decades, but he says "At the same time, 
theropods are referred to as nonavian theropods. This implies (but has never 
been said) that birds (= Avialae) are avian theropods, feathered dinosaurs, 
or living dinosaurs (Gauthier and Gall, 2001)."  Surely the 'birds are 
dinosaurs' slogan has been widespread, enough to inspire DML threads by 
those annoyed by it.

Kurochkin notes the numerous feathered coelurosaurs now known, also 
crediting Early Jurassic theropods, pterosaurs and Psittacosaurus with 
sufficiently similar structures (the former is based on Gierlinski's (1996) 
tracks which preserve vegetation mats, not feathers).  He suggests that 
because filamentous integument was so widespread, it should not be assumed 
to be homologous in coelurosaurs and birds.  What he ignores is the detailed 
similarity between microraptorian and ornithurine stage III feathers, for 
instance.  Even if Psittacosaurus' quills and pterofuzz are homologous to 
feathers, they lack barbs and a central rachis, making them stage I in 
Brush's scheme.

The author correctly notes some characters are convergently present in 
ornithurines and some maniraptoran lineages, such as oviraptorids' 
double-headed quadrate and Nomingia's pygostyle.  Yet these and other 
attempts to dehomologize characters (uncinate processes, opisthopuby) aren't 
defended rationally.  The fact structures serve different purposes in 
different taxa does not mean they lack homology, nor does differing 
morphology indicate lack of homology.  As an obvious example, whale tails 
have a different purpose and morphology than cow tails, yet both are 
homologous.

Both of these paragraphs illustrate a fallacy of Kurochkin's- that if a 
structure is found to be homoplasious in two taxa, its homology in a third 
taxon can be safely doubted.  So because pygostyles developed in Nomingia 
and ornithurines convergently, it's easier to assume the pygostyles of 
enantiornithines and ornithurines are convergent.  Yet evolutionary lineages 
are independent of one another, and no character is a priori immune to 
homoplasy.

Kurochkin discusses the I-II-III II-III-IV disparity between basal theropods 
and neornithines, and notes that digital frameshifts have been shown to 
occur in living taxa.  He then falls into the classic BAND trap of naive 
falsification (Makovicky and Dyke, 2001)- "The supporters of the dinosaurian 
hypothesis for the origin of birds believe that this is an open question 
that requires further examination. However, the recognition of distinctions 
between theropods and birds in the manual formula alone conflicts with the 
sisterly or ancestor?descendant relations of these groups."  And even more 
explicitly- "To conclude the consideration of the major characters shared by 
theropod dinosaurs, Archaeopteryx, and birds, note that, as relationships 
are estimated, cladistics do not take into account differences; nonetheless, 
we believe that they may be of primary significance for this purpose. Such 
differences unequivocally indicate the absence of direct relationships of 
living ornithurine birds with predatory dinosaurs."  Kurochkin states the 
ornithurine tarsometatarsus could not evolve from an arctometatarsus, but no 
BADist claims it did.  Oddly, he says heterocoelous vertebrae cannot evolve 
from opisthocoelous ones, though several enantiornithines have partial 
heterocoely, as do some deinonychosaurs.  Even stranger, he seems to think 
birds' single oviduct couldn't evolve from one of a theropod's two oviducts, 
but as far as I can tell amniotes primitively have two oviducts.  So unless 
birds aren't amniotes, they had to lose an oviduct somewhere.  Finally, he 
believes theropods had a diaphragm giving them a fundamentally different 
respiratory mechanic than birds, despite the numerous objections raised by 
Paul et al. (immobile pubis, etc.) and fundamentally birdlike thoracic 
morphology of maniraptorans including enantiornithines (airsacs, uncinate 
processes, large sterna, ossified sternal ribs, etc.).

Kurochkin's cladogram is-
Archosauromorpha
|--Theropoda
|..|--Coelurosauria
|..`--+--Troodontidae
|.....`--+--Alvarezsauridae
|........`--+--Oviraptoridae
|...........`--+--Dromaeosauridae
|..............`--Sauriurae
|.................|--Vorona
|.................`--+--Archaeornithes
|....................`--Enantiornithes
`--+--Protoavidae
...`--+--Confuciusornithidae
......`--Ornithurae
.........|--Zhyraornithidae
.........`--+--Liaoningornis
............`--+--Ichthyornithes
...............`--+--Hesperornithes
..................`--Neornithes
However, his phylogram shows Vorona being derived from enantiornithines.  It 
also indicates he views Wyleyia, Horezmavis, Gargantuavis, patagopterygids 
and kuzholiids as ornithurines.

Discussing Archaeopteryx, Kurochkin correctly notes most of its birdlike 
characters have been found in other theropods, though many of his examples 
are subtly flawed.  For instance, he states one such character is "the 
double-condyled quadrate (although it is single-condyled in Archaeopteryx 
and Enantiornithes, while in Ornithurae it is double-condyled)."  In 
actuality, most coelurosaurs (including tyrannosaurids, caenagnathoids, 
troodontids and Archaeopteryx) have a surface on the quadrate which contacts 
the braincase.  Most basal euornithines are similar, an in fact the only 
taxa with a distinct double-condyled morphology are Shuvuuia, 
Confuciusornis, Enaliornis and many neoavians.  Euornithines with basically 
single-headed quadrates include Archaeorhynchus, Patagopteryx, Yixianornis, 
Apsaravis, Potamornis, Hesperornis, Ichthyornis, paleognaths and 
galloanserines.  Another example- "the fusion of the proximal tarsals with 
the tibia and fibula, at least in adults (they fuse in the majority of adult 
theropods; however, in ornithurine birds, the fibula does not participate in 
this fusion)."  These do not in fact fuse in most adult theropods.  Nor do 
most enantiornithines leave their distal tarsals unfused, contra Kurochkin.  
He goes on to list several characters more recently proposed to unite 
Archaeopteryx with birds (external naris extends far posteriorly, few caudal 
vertebrae, reduced postacetabular process, etc.), but confusingly says "The 
artificiality of placing Archaeopteryx close to Ornithurae based on such 
characters does not need any proof."  Well Kurochkin, yes it does.  Finally, 
Kurochkin lists 16 "derived" characters Archaeopteryx shares with theropods 
but not with ornithurines.  Symplesiomorphies, of course (how can he 
possibly think the presence of an ectopterygoid or long pubic symphysis is 
DERIVED within Reptilia?!).  Taking each in turn-
(1) Wide dorsal end of the lacrymal. Also in Archaeorhynchus.
(2) The anteriorly directed medial condyle of the ventral portion of the 
quadrate. Also in Confuciusornis and Hesperornis.
(3) The presence of the ectopterygoid. Also in Confuciusornis.
(4) Uniform width of the body of the scapula throughout its extent. Also in 
Confuciusornis and Hesperornis.  Absent in enantiornithines.
(5) Flat subquadrate coracoid. Absent in enantiornithines.
(6) Special tubercle located cranioventral to the glenoid on the coracoid.  
Also in Patagopteryx and Hesperornis.  Modified into the acrocoracoid in 
enantiornithines, as in confuciusornithids and ornithurines.
(7) The projection of the plane of the proximal head of the humerus is 
directed perpendicular to the distal head.  Also in Confuciusornis. Absent 
in Neuquenornis and many theropods.
(8) The forearm is shorter than the manus and humerus. Also in 
confuciusornithids, Hongshanornis, Yixianornis.  Absent in many 
enantiornithines.
(9) Digits IV and V of the manus are reduced.  Only absent in 
ornithuromorphs according to Kurochkin due to his disbelief in the 
frameshift.
(10) The semilunate carpal is articulated with metacarpals 1 and 2.  Ditto.
(11) The ungual phalanges of the manus are large, with large tubercles for 
the insertion of flexor tendons.  Also in confuciusornithids.  Absent in 
enantiornithines.
(12) Development of the lesser trochanter in the proximal part of the femur. 
 Absent in enantiornithines.
(13) Long preacetabular part of the ilium compared to the short 
postacetabular part.  Also in Confuciusornis, Archaeorhynchus, and 
Patagopteryx.  Absent in Aberratiodontus.
(14) Development of a peduncle on the ilium for the articulation with the 
pubis.  Also in Confuciusornis, Archaeorhynchus, Patagopteryx, Apsaravis, 
Gansus, Ichthyornis and Iaceornis.
(15) Split of the caudal end of the ischium.  Absent in enantiornithines.
(16) The pubis has a large symphysis.  Also in confuciusornithids, 
Archaeorhynchus, Hongshanornis, Yanornis and Yixianornis.
Remember that in Kurochkin's view, confuciusornithids are on the bird branch 
while enantiornithines are on the theropod branch.  So of the 14 non 
frameshift-influenced "theropod" characters above, confuciusornithids have 
nine.  At least some enantiornithines lack nine of them.  Of the remaining 
five "ornithurine" characters, four are missing in at least basal 
euornithines.  The remaining character involves the ectopterygoid's absence, 
which has not been verified in any ornithurine more basal than Hesperornis.  
If confuciusornithids (as basal birds) have most theropod synapomorphies, 
why not just have birds derive from theropods?  And if enantiornithines can 
lose theropod synapomorphies, why not let birds lose them as well?

Kurochkin believes many flight-related characters in enantiornithines and 
ornithurines arose through convergence, a conclusion I'm increasingly 
agreeing with as taxa like Hongshanornis and Archaeorhynchus are described.  
This does not mean every ornithothoracine character is convergent however.  
Kurochkin uses symplesiomorphies to group enantiornithines with 
Archaeopteryx instead-
1. "the quadrate has a laterocaudal condyle;"  Seen in all Mesozoic 
theropods, including Confuciusornis and Hesperornis.
2. "the quadratojugal and quadrate are articulated by the adjacent joint 
(Martin and Zhou, 1997);" As in Confuciusornis and Hesperornis.  How else 
would they articulate?
3. "the distal head of the femur has a lateral crest;"  Absent in 
Archaeopteryx (Chiappe, 2002).
4. "the proximal ends of the metatarsals are arranged tightly to form an 
integral transverse row."  I honestly don't understand this character, 
though it seems to be a standard in the sauriurine lists.  All theropod 
metatarsals are tightly arranged in a mesotarsal manner.  Perhaps he means 
metatarsal III is not displaced plantarily or that there is no intercotylar 
eminence, but this is true of Archaeorhynchus too, not to mention 
confuciusornithids.  Indeed, there's really no metatarsal feature 
Archaeopteryx and enantiornithines share to the exclusion of 
confuciusornithids.
Finally, Kurochkin mentions the LAGs found in enantiornithines, and 
indeterminate growth of the latter(?) and Archaeopteryx.  Both are 
archosaurian plesiomorphies that cannot support Sauriurae however.  As no 
valid characters support Sauriurae, the ornithothoracine characters which 
are seen even in basal enantiornithines and euornithines (pygostyle, 
strut-like coracoid, trochanteric crest, etc.) may be used to construct the 
phylogeny.  Interestingly, the addition of basal euornithines has lowered 
the support for Ornithothoraces in Clarke's matrix recently (Zhou and Zhang, 
2005, 2006), but that groups confuciusornithids and enantiornithines 
together, in contrast to Kurochkin's phylogeny.

Kurochkin places confuciusornithids on the ornithurine line in his 
phylogeny, seemingly due mostly to the absence of LAGs.  As seen above, it 
would make much more sense for him to place them in his theropod lineage.

Kurochkin lists 18 characters for his Ornithurae (= Ornithuromorpha or 
Euornithes; excludes confuciusornithids and Protoavis).  Of them, several 
are missing in basal taxa such as Archaeorhynchus and Hongshanornis 
(heterocoelous cervical vertebrae; at least ten sacral vertebrae; gastralia 
reduced; elongate sternum; pubis parallel to ilium; intercotylar eminence on 
metatarsus; metatarsal III plantarily displaced; completely fused 
metatarsus), while others are present in some enantiornithines (more than 
nine cervical vertebrae; elongate sternum; capital groove on humerus; broad 
flat manual phalanx II-1; distal tarsals fused to metatarsus; metatarsals 
completely fused).

Interestingly (and disasterously), Kurochkin concludes with Protoavis.  He 
places this taxon as the basalmost of his bird lineage, basal to 
confuciusornithids and ornithurines.  This is based on several characters-
1. heterocoelous cervical vertebrae.  If taken to mean completely 
heterocoelous vertebrae (as to distinguish neornithines from 
enantiornithines), absent in Confuciusornis, Archaeorhynchus, Yixianornis, 
Gansus and Ichthyornis.
2. voluminous tropibasal braincase.  I don't know what tropibasal means, 
though no non-ornithurine Mesozoic bird braincases besides Archaeopteryx 
have ever been measured volumetrically.
3. reduced postorbital.  Absent in Confuciusornis.
4. closed temporal fenestrae.  Absent in Confuciusornis, correlated with 
previous character.
5. narrow elongated coracoid.  Also in enantiornithines
6. mobile articulation of the scapula to the coracoid through a fossa in the 
coracoid and a prominence on the scapula.  Absent in confuciusornithids.
7. deep renal depressions on the internal surface of the ilium and sacrum.  
Absent in Confuciusornis, Patagopteryx, Yixianornis, Apsaravis, 
hesperornithines and Ichthyornis.
8. fusion between both ends of the third and fourth metacarpals.  Absent in 
confuciusornithids, Archaeorhynchus and Hongshanornis.
9. plantar shift of the third metatarsal in the proximal part of the foot.  
Absent in confuciusornithids, Archaeorhynchus and Patagopteryx.
It's amazing this escaped Kurochkin's notice (even accepting Protoavis as a 
distinct taxon instead of a chimaera).  The near complete lack of these 
characters in confuciusornithids should suggest Protoavis is more closely 
related to ornithurines, or that it developed the characters convergently 
with ornithurines.  Indeed, the renal fossa is a much better example of 
homplasy than anything Kurochkin listed for theropods vs. birds.  Not only 
does Confuciusornis lack it, but _all known non-neornithine birds_ do too, 
when they can be examined.  The absence of heterocoelous cervicals, a renal 
fossa, distal metacarpal fusion and a plantarily displaced metatarsal III in 
basal ornithurines invalidates their use as synapomorphies for Protoavis + 
Ornithurae.  This leaves the scapulocoracoid joint, strut-like coracoid, 
enlarged braincase and reduced postorbital as potentially valid 
synapomorphies, though enantiornithines share the second and possibly third, 
while the third and fourth are unconfirmed in basal ornithurines.  Also 
notable is that Protoavis has a few of Kurochkin's supposed theropod 
characters- anteriorly directed medial quadrate condyle, lesser trochanter 
on femur, pubic peduncle on ilium.
Kurochkin claims "Late Triassic Protoavis is a headache for the proponents 
of the dinosaurian hypothesis of the origin of birds. If its characters were 
introduced in a cladistic matrix, all cladograms of sister relationships of 
theropods and birds would be destroyed, while the Late Jurassic dating for 
the deviation of birds from the dinosaurian stem would be rejected; the same 
is true of the differentiation of the basic theropod families, which would 
be shifted to the Late Triassic. This is probably the main reason why the 
majority of supporters of sister relationships between birds and theropods 
omit Protoavis from their considerations."  No, the main reason is that the 
remains are identical to coelophysoid and pterosauromorph (Atanassov, 2002) 
elements, while the cervicals are similar to drepanosaurids'.  Besides being 
a chimaera, many details described and illustrated by Chatterjee are not 
recognizable in the fossils (Witmer, 2001).  For instance, the metacarpus 
Kurochkin uses to propose a character uniting Protoavis with ornithurines is 
actually a pterosauromorph metatarsus.  While the cervicals he uses are 
extremely similar to contemporary drepanosaurids.  Admittedly, the 
apparently reduced postorbital and bone resembling a euornithine coracoid 
are intriguing.  But a massive reanalysis involving comparison with various 
Triassic vertebrate elements is needed before we'll know which, if any, 
material belongs to a distinct taxon, let alone a bird.

On the one hand, Kurochkin appears to have a much better grasp of BAD than 
Feduccia, Martin or Czerkas.  Yet he is just as guilty of naive 
falsification and magically knowing which characters are free of homoplasy 
and thus important for phylogeny.  Also similar to Martin is the presence of 
large amounts of phylogenetic contradictions which remain undiscussed.  
Martin (2004) proposed Protoavis as a basal bird, ignoring the fact it 
possessed his supposed theropod characters, lacked characters he claimed to 
be primitive for birds, and lacked characters he claimed (the supposedly 
more basal) Longisquama shared with birds.  Equivalently, Kurochkin proposes 
confuciusornithids as basal birds, ignoring the fact they possess most of 
his supposed theropod characters, and lack characters he claimed (the 
supposedly more basal) Protoavis shares with birds.  Kurochin's theropod + 
Archaeopteryx characters are mostly present in confuciusornithids, and 
mostlyly absent in enantiornithines, with several seen in basal 
euornithines.  Just what's to be expected if the standard topology were 
true.  Kurochkin's sauriurine characters are few and symplesiomorphic, as 
Martin's and Feduccia's have always been.  Kurochkin's placement of 
confuciusornithids as more closely related to ornithurines than 
enantiornithines contradicts nearly all of his own characters, let alone 
those of Clarke, Chiappe and others.  His placement of Protoavis in this 
position is weakened by the absence of half the states in basal ornithurines 
and the presence of a few theropod characters in it.  Kurochkin's general 
point that avian convergence is rampant in coelurosaurs is quite valid, 
though his functional/morphological method of showing this is unsound.  
Similarily, he's right that Archaeopteryx is not necessarily more birdlike 
than some traditionally more basal coelurosaurs, but his utterly random 
rejection of the birdlike characters it DOES have only hurts his argument.  
And while many ornithothoracine characters may have been convergent in 
derived enantiornithines and euornithines, it's the basal members of these 
clades lacking the characters which shows this is true, not homoplasy of the 
character in outgroups or differing morphologies in the ingroups.  The 
methodological and logical problems noted above prevent Kurochkin's model 
from being a likely alternative to BAD.

Mickey Mortimer