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

Confuciusornithid Taxonomy

Dan Benson just asked about this, so I figured I'd post what I've had
written for a while now....

Hou et al. (1999) described Confuciusornis dui based on two skeletons, the
holotype complete (IVPP V11553) and the paratype missing the forelimbs and
skull (IVPP V11521).
Confuciusornis dui shows the following differences from C. sanctus.
1. narrow, tapered and vertical ascending process of maxilla without
maxillary fenestra
2. maxilla takes up most of ventral orbital margin
3. low rounded dorsal jugal process?
4. jugal forms ventral margin of laterotemporal fenestra?
5. mandible without anteroventral expansion
6. posterodorsal dentary process extends posteriorly over most of mandibular
7. posterodorsal dentary process taller than posteroventral process
8. no ventral surangular process invading mandibular fenestra
9. reduced surangular foramen
10. sternum with anterior notch
11. sternal ribs attach to lateral processes
12. sternal lateral processes not bifurcate
13. four sternal ribs
14. sternal ribs grow posteriorly shorter, so that last is less than a third
as long as the first
15. last two sternal ribs markedly expanded distally
16. metacarpal I tapers proximally
17. manual unguals I and III subequal in size.
Hou et al. (1999) identified characters 5, 10 and 17.  They also proposed
the following apomorphies- mandible more slender anteriorly; beak more
pointed anteriorly; sternum more elongate; lateral processes on sternum;
tarsometarsus shorter than pygostyle.  The first three are not noticeably
different from C. sanctus, while the last two are found in C. sanctus.
Changchengornis lacks 1, 16 and 17, but has 9.  Characters 1, 2, 3, 6, 7,
11, 14, 15, 16 and 17 are apomorphies of C. dui.  The opposite of characters
5, 8, 9 and 13 are apomorphies of C. sanctus.  The polarity of characters 4,
10 and 12  is uncertain.  Confuciusornis sanctus actually shares the
opposite of 10 and 12 with more derived pygostylians.  So these could either
be independantly developed in C. sanctus and other pygostylians, or lost in
C. dui.

Hou (1997) described Confuciusornis chuonzhous, based on a distal
tibiotarsus and pes (IVPP V10919) which was originally a paratype of C.
sanctus.  Hou lists the following characters in its diagnosis-
1. large and robust
The metatarsus is 33 mm long, which is comparable to some C. sanctus
specimens (GMV-2132, GMV-2154, IVPP V110304).  It actually appears more
gracile than C. sanctus specimens.  Hou cites the tibiotarsal width as 4 mm
and the metatarsal length of 33 mm, compared to the C. sanctus paratype's
2.5 mm and 22 mm respectively.  The ratios of 12.1 vs. 11.4 are not
significant in my opinion.
2. tibiotarsus anteroposteriorly thick
No evidence is given, but would be hard to determine in a specimen crushed
in anterior view.
3. distal tibiotarsus unexpanded
This is untrue.
4. distal condyles indistinct and not projected anteriorly
Not apparent from the photo or figure, but could be due to the specimen's
poor preservation.
5. astragalus and calcaneum separate from tibia
Same as above, as noted by Chiappe et al., who also pointed out such a
difference would be more likely ontogenetic.
6. pedal digit I with three phalanges
Hou interpreted metatarsal I as pedal phalanx I-1, and phalanx I-1 as I-2.
7. pedal ungual I smaller
8. pedal ungual I with reduced curvature
These two characters are probably caused by the ungual being broken
9. metatarsal V larger
C. sanctus has a fifth metatarsal 36% of metatarsal IV's length (GMV-2133),
while C. chuonzhous' is 32%, assuming the metatarsal next to V is IV.  This
is uncertain because of the specimens poor preservation.  Digits I and V are
on the same side and digit IV is seemingly articulated with the shortest
main metatarsal (would normally be II).  If I have the metatarsals
backwards, V would be even shorter.
This specimen is very poorly preserved, and therefore difficult to evaluate.
Looking at the differences between Confuciusornis and Changchengornis below,
Hou's illustration of C. chuonzous agrees with Confuciusornis in not having
distal fusion between metatarsals III and IV, and in having a slender
trochlea on IV.  It must be mentioned that Hou's illustrations are not the
most accurate, and I could not confirm these characters from a photo of the
fossil.  No hindlimb characters have been found so far to separate C.
sanctus and C. dui.  In conclusion, I suggest this specimen be referred to
as Confuciusornis? sp. indet..

Hou (1997) described Confuciusornis suniae, based on a nearly complete
skeleton (IVPP V11308).  Unfortunately, it is not illustrated, but a photo
is available.  One is online here-
He listed the following characters in the diagnosis of Confuciusornis
1. V-shaped anteromedian notch in premaxilla
This is present in C. sanctus (GMV-2133).
2. longer nasal process of premaxilla
This is due to Hou interpreting the posterior portion of C. sanctus' nasal
process as the nasals and anterior frontals.
3. external naris large
This is due to Hou interpreting the external nares of C. sanctus as
antorbital fenestrae.
4. frontals short
This is due to the same error as 2.
5. parietals developed
This is too vague to be useful.
6. cervical vertebrae broad
They do not appear any more so than in C. sanctus, and Hou acknowledges
there may not be a difference.
7. cervical pleurocoels present
This is present in C. sanctus (Jm-UKr-1997/1) and C. dui, but is not visible
in most specimens due to poor preservation.
8. cervical neural spines narrow and low
This is true in C. sanctus as well.
9. dorsal vertebrae long and thin
This is too vague to comment on.
10. long deep grooves in dorsal pleurocoels
The interior morphology of C. sanctus' dorsal pleurocoels (fossae?) is
11. last three dorsal diapophyses and neural spines fused
This is due to Hou interpreting the first three sacrals as the last three
12. sacral vertebrae fused to ilia
This is not confirmed to be absent in C. sanctus and would be expected to
vary ontogenetically in any case.
13. sacral neural spines fused
This is present in C. sanctus.
14. about fifteen caudals fused into a pygostyle
A pygostyle formed of many caudals is also present in C. sanctus.
15. sternum "narrow and long with deep elongated lateral recesses"
Chiappe et al. (1999) confirm the sternal morphology described by Hou cannot
be verified.
In addition, he cited the following supposedly diagnostic features in the
16. bulbous frontal with thickened orbital rim
Chiappe et al. confirm the first is present in C. sanctus and the second
isn't known to be absent, but is common in coelurosaurs.  It is possible Hou
confused the palpebrals for an orbital rim though.
17. lateral process on cervical diapophyses
These are not described well, but may be transverse processes.
18. cervical prezygopophyses "positioned anterolaterally to centrum and
projected anteriorly"
This morphology is seen in C. sanctus as well.
19. thin-(walled?) cervical centra
If interpreted correctly, the condition is common in theropods.
20. sternum heart-shaped
21. sternum lacking lateral processes
Both indeterminate, as noted above.
22. proximal humeral foramen smaller
Chiappe et al. note it is within the range of variation of C. sanctus.  I
see no difference between it an other specimens.
23. less prominent ventral tubercle of humerus
C. sanctus lacks a ventral tubercle (Chiappe et al., 1999), so this
character is invalid.
24. dorsal supracondylar process on distal humerus
If true, this would differ from C. sanctus, but I see no evidence of its
25. low and concave dorsal epicondyle of humerus
Chiappe et al. describe a round fossa that excavates the dorsal margin of
the distal humerus in C. sanctus.
26. scapulotricipital sulcus on distal humerus
This is not confirmed to be absent in C. sanctus.
27. metacarpal III more robust
I observe no difference between the specimen and C. sanctus.
28. ischium with ilial peduncle, proximodorsal process and shaft
Hou describes the ischium as so- "The ischium is extremely autapomorphic as
there are three processes extending off the main body: The first is
relatively short, broad, and forms the posterior wall of the acetabulum. The
second process is plate-shaped, extends, and expands obliquely dorsally from
the distomedial side of the ilium to the vertebral column, appearing as
though it encircles the most posterior portion of the synsacral vertebrae.
The third and largest process is one which extends posteriorly to the distal
end."  The first process is clearly the ilial peduncle, the second is the
proximodorsal process, and the third is the ischial shaft itself.  These are
found in C. sanctus too.
29. prominent fourth trochanter
I observe only a slight swelling, presumedly caused by crushing.
30. prominent trochlea fibularis on distal femur
Chiappe et al. confirm this is present in C. sanctus as well.
31. prominent fibular crest on tibia
This is also present in C. sanctus.
32. metatarsus only fused proximally
As seen in C. sanctus.
33. robust limb bones
Cannot be confirmed, though if true perhaps due to the crushing that has
affected most of the elements.
Hou also identified a septomaxilla in the anterior corner of the external
nares.  This element is of course absent in birds, and similar ambiguous
ossifications are seen in some C. sanctus specimens (eg. GMV-2132).  These
may be palatal elements, or perhaps ossified nasal airway walls.  A dorsal
articular process like that of dromaeosaurids is described.  Nine cervicals
are preserved, verifying at least this many were present in life (specimens
examined by Chiappe et al. had seven or less).  Ten free caudals are
preserved, and the vertebrae making up the pygostyle are distinct, though
the neural spines are fused distally.  This is distinct from C. sanctus,
which has seven free caudals and no differentiation within the pygostyle.
Perhaps the specimen is young and the two distalmost free caudals would fuse
into the pygostyle when it matured.  Hou describes four "short thick rib
segments" on the right of the specimen, which I think may be uncinate
processes.  He also states that though an olecranal fossa is absent, a small
depressed region is observed, whereas Chiappe et al. noted no sign of an
olecranal fossa.  Although Chiappe et al. state a longitudinal groove is
absent on the radius of C. sanctus, Hou reports such a groove in C. suniae
that is present proximally.  Hou reports five phalanges on manual digit III,
the first two very short.  This is near certainly caused by a break near the
base of phalanx III-2.  The ilial process noted over the acetabulum and
compared to an antitrochanter is probably the supratrochanteric process.  He
notes a "low dorsal ridge" on the distal pubis, similar to that described by
Paul (2002).  A patella is described, though this has not been reported by
Chiappe et al., and I am unaware of its presence in any non-ornithurine
bird.  Hou claims the notched premaxillary tip, subequal length of pedal
digits III and IV and enlarged fourth pedal ungual indicate semiaquatic or
aquatic habits.  I see no reason these characters suggest this, and the
first two are present in C. sanctus as well.  As reported, the fourth pedal
ungual is 20% longer than the third, which is unlike at least some C.
sanctus specimens.  However, pedal proportions among confuciusornithids are
extremely variable, with digit length differing as much as 30% in feet of
the same specimen.
The proximal humeral foramen and bowed manual phalanx II-2 confirm this is a
Confuciusornis specimen, while the elongate sternal ribs, proximally robust
first metacarpal and reduced manual ungual III suggest it is not C. dui.
The ten free caudals, slightly developed olecranal fossa, radial groove and
patella all differ from C. sanctus, but given Hou's record of erroneous
interpretation, I'm cautious to accept these as real.  I thus provisionally
agree with the synonymization of Confuciusornis suniae with C. sanctus, as
suggested by Chiappe et al..

Changchengornis shows the following differences from Confuciusornis sanctus-
1. decurved premaxilla
2. rostrum <40% of cranial length
3. surangular rises sharply at posterior edge of external mandibular
4. small surangular foramen
5. furcular arm width <20% of length
6. tubercle on center of posterior furcula
7. furcula grooved anteriorly and posteriorly
8. humerus subequal to femur in length
9. no proximal humeral foramen
10. metacarpal I >40% of metacarpal II
11. manual phalanx II-2 straight
12. manual phalanges II-1 and II-2 subequal in length
13. metatarsals III and IV fuse distally
14. trochlea of metatarsal IV subequal in width to metatarsals II and III
15. tarsometatarsus not excavated on plantar surface
Chiappe et al. (1999) identified characters 1, 2, 3, 6, 7, 9, 10, 12 and 13
in their diagnosis.  They also proposed the following apomorphies- mandible
much shorter than skull; posteromedian process of sternum narrower; hallux
longer compared to digit II.  The first cannot be confirmed because the
anterior dentary does not connect to the posterior  mandible in the
holotype.  The second and third are within the range of variation seen in
Confuciusornis (IVPP V11521 and IVPP 11308 respectively).  Confuciusornis
dui lacks 1, 2, 3, 8, 9, 10, 11, 12, 13 and 14, but has 4.  Characters 1, 2,
3, 6, 7, 8, 10, 12 and 13 are apomorphies of Changchengornis.  The opposite
of characters 4, 5 and 15 are apomorphies of Confuciusornis sanctus.  The
opposite of characters 9, 11 and 14 are apomorphies of Confuciusornis.

As for Hou et al.'s (2001) new genus Jinzhouornis, it's definitely
confuciusornithid, but I have yet to see its description.
At least some of the characters diagnosing J. yixianensis are seen in
Confuciusornis (eg. snout over 50% of cranial length; tubercle in middle of
metatarsal II; small tubercle on metatarsal III) and most others are vague
proportions that could also be applied to C. sanctus specimens (eg. long low
skull; robust long snout; moderately sized orbit; extremely curved manual
unguals; slender humerus).  Characters like "braincase small", "second
manual digit not particularly expanded" and "scapula about as long as
humerus" appear to differ at first glance, but I have a feeling examination
of the specimen would show otherwise.  For instance, the humerus is clearly
broken and was much longer than the scapula if complete.  Also, I cannot see
digit II, but I do see digit III, which is of course more slender.  Better
photos would be needed to show the cervical vertebrae are shorter than
Confuciusornis, and Confuciusornis may have had over twelve dorsal
vertebrae.  The taxon shows the proximal humeral foramen of Confuciusornis,
but cannot be distinguished from C. sanctus or C. dui based on available
information.  I provisionally recommend it be synonymized with
Confuciusornis, perhaps as C. yixianensis, until a more detailed study can
be performed.
J. zhangjiyingia is similarly doubtful.  The diagnosis contains characters
that could be applied to Confuciusornis sanctus (eg. long large skull;
premaxilla extends to posterior part of orbit; infratemporal fenestra well
developed; orbit moderate in size; furcular ends widely separated).  Others
cannot be confirmed (eg. nasals lie ventral to frontals; quadratojugal
contacts orbit) or are subject to individual variation (humeral shaft
robust; furcula slender).  It shows the bowed manual phalanx II-2 and
humeral foramen of Confuciusornis.  The first manual ungual is much larger
than the third, like C. sanctus.  I provisionally refer Jinzhouornis
zhangjiyingia to Confuciusornis sanctus pending restudy.

Note I haven't compared the probable confuciusornithid "Proornis".

So, using only derived characters-

Diagnosis- proximal humeral foramen; manual phalanx II-2 bowed; trochlea of
metatarsal IV much smaller than those of metatarsals II and III.

Diagnosis- anterior dentary expanded ventrally; ventral surangular process
invading external mandibular fenestra; enlarged surangular foramen; furcular
arm width >20% of arm length; five pairs of sternal ribs; tarsometatarsus
excavated on plantar surface.

C. dui
Diagnosis- narrow, tapered and vertical ascending process of maxilla without
maxillary fenestra; maxilla takes up most of ventral orbital margin; low
rounded dorsal jugal process?; jugal forms ventral margin of laterotemporal
fenestra?; posterodorsal dentary process extends posteriorly over most of
mandibular fenestra; posterodorsal dentary process taller than
posteroventral process; sternum with anterior notch; sternal lateral
processes not bifurcate; sternal ribs attach to lateral processes; sternal
ribs grow posteriorly shorter, so that last is less than a third as long as
the first; last two sternal ribs markedly expanded distally; metacarpal I
tapers proximally; manual unguals I and III subequal in size.

Changchengornis hengdaoziensis
Diagnosis- decurved premaxilla; rostrum <40% of cranial length; surangular
rises sharply at posterior edge of external mandibular fenestra; tubercle on
center of posterior furcula; furcula grooved anteriorly and posteriorly;
humerus subequal to femur in length; metacarpal I >40% of metacarpal II;
manual phalanges II-1 and II-2 subequal in length; metatarsals III and IV
fuse distally.

Mickey Mortimer