|
Microraptor Xu, Zhou and Wang 2000
M. zhaoianus Xu, Zhou and Wang 2000
Etymology- "Zhao's small thief", after renouned
paleontologist Zhao Xijin.
Barremian, Early Cretaceous
Yixian Formation, Liaoning, China
Holotype- (IVPP V 12330) (48 cm) partial skull (45
mm), lower jaws, four dorsal vertebrae, several dorsal ribs, sternal ribs or
uncinate processes, gastralia, sacrum (19 mm), caudal series (24-25 vertebrae)
(246 mm), chevrons, radius, ulna (35 mm), radiale, ulnare, semilunate carpal,
fourth carpal, proximal metacarpus, partial ilia, incomplete pubes, ischia,
femora (one fragmentary) (53 mm), tibiae (one partial) (68 mm), fibula,
metatarsi, pedes, feather impressions, keratin claw sheaths
Diagnosis- only posterior teeth have constricted
roots; accessory crest on femur; very slender ischium; obturator process whose
tip extends further distally than the dorsodistal corner of the ischium;
near-arctometatarsalian metatarsus; metatarsal IV most robust.
Description-
This is the "dromaeosaur" part of Archaeoraptor
liaoningensis. The other part is a pygostylian that remains to be
described. This species is claimed to be the smallest non-avian theropod,
a distinction it shares with Parvicursor. Using Archaeopteryx as a guide,
but adjusting for the longer tail of Microraptor (both had similarily
proportioned heads, presacral vertebrae are poorly known for Microraptor), gives
an estimate of 48 cm for its total length. Several features attest to its
maturity, including the fused sacrals, pubic symphysis, last dorsal partially
fused to the sacrum and well-ossified cortical bone.
The skull is badly preserved, except for the
snout. The premaxilla is intermediate between the pointed rostrum of
Archaeopteryx and Sinornithosaurus and the blunt snout of Deinonychus and
Velociraptor. The subnarial process is shorter than any of these,
suggesting the naris contacted the maxilla. The external naris itself was
placed in a position similar to dromaeosaurids, not further back as in
Archaeopteryx. Premaxillae were unfused. The maxilla and nasals show
no useful information, except that the top of the snout was fairly straight and
the external naris did not extend posteriorly much past the subnarial
process. The only other identifiable cranial element is the quadrate,
which shows no useful details. The lower jaws are preserved, although
distorted. The dentary is slender and straight laterally, but bows
medially in dorsal view. There are at least fifteen teeth, which are
placed in separate alveoli. The posterior lower jaw appears to be
preserved in medial(?) view, as the internal mandibular fenestra is shown.
It is said to be different than the large, triangular fenestra of
dromaeosaurids. Teeth vary with position. The premaxillary and
anterior dentary teeth are flattened, recurved and lack serrations. Their
roots are of normal theropod type. Posterior teeth possess posterior
serrations, but no anterior serrations. Their crowns are not as compressed
and they have constricted roots. Serrations are not enlarged, numbering
13-14 per mm on a 2 mm tooth crown.
The four preserved dorsal vertebrae are apparently
from the anterior end of the series, as they possess large hypapophyses.
Structures that are either sternal ribs or uncinate processes are preserved with
the dorsal ribs. Gastralia are also present. The sacrum contains
five incompletely fused vertebrae, the first three of which are expanded
transversely. The ventral surfaces of the first three seem to lack
prominent grooves or keels. Twenty-four or twenty-five caudal vertebrae
are present. Transverse processes end before caudal eleven, but most
caudals possess dromaeosaurid-like elongate prezygopophyses and chevrons.
The distal caudal centra are elongate (to at least 130% of proximal centrum
length).
The ulna is strongly bowed and the radius is less
than 70% of ulnar width. The ulnofemoral ratio is .66, which is comparable
to dromaeosaurids, Protarchaeopteryx and some oviraptorosaurs, while it is much
smaller than Archaeopteryx and other volant forms (>.8). This suggests
Microraptor was flightless. Four carpals are present, including the
semilunate. The semilunate carpal has a larger contact with metacarpal II
than metacarpal I, which differs from dromaeosaurids, but agrees with avians and
more basal taxa. The unfused bases of three metacarpals are preserved,
with mc II being most robust and mc III being least.
Only portions of the acetabular area and the
postacetabular processes are visible in the ilia, but these don't reveal much
information. The supracetabular crest is probably absent and the
postacetabular processes flare outward and taper sharply, suggesting a reduced
brevis shelf. The pubes are preserved in anterior view. The authors code
this taxon as having a mesopubic pelvis (similar to Unenlagia, Rahonavis and
Archaeopteryx). The symphysis is 49% of pubic length and the pubic apron
is wider than dromaeosaurids. It tapers abruptly following the symphysis
and has parallel sides further distally until it ends bluntly. The ischium
is merely 47% of pubic length, but is much more slender than other
eumaniraptorans. There is a narrow, triangular obturator process whose tip
extends further distally than the dorsodistal corner of the ischium.
A mediodorsal process is present, but there is no proximodorsal
process. Despite its slenderness, the ischium is said to be plate-like in
section.
An accessory trochantor is present at the base of
the lesser trochantor. The tibia is 1.28 times femoral length and the
fibula reaches the calcaneum. The metatarsus is unfused and subequal to
the radius in length. Metatarsal I is reversed and placed distally.
The second digit has phalanx II-2 >85% of phalanx II-1, as in dromaeosaurids
and avians. Although the ungual is enlarged and possesses a larger flexor
tubercle than the other unguals, phalanx II-2 is only slightly expanded
proximally. It lacks the proximoventral heel of dromaeosaurids and
troodontids, and the proximoventral tubercle of Rahonavis. Metatarsal III
tapers proximally to the point that it may be arctometatarsalian.
Metatarsal IV is the most robust, similar to troodontids, but exhibits a
posteromedial flange similar to dromaeosaurids. The fifth metatarsal is
elongate and laterally bowed.
The feathers are preserved around the ulna, manus,
pelvis, femur and tibia. They have an outline similar to those of Yixian
birds. Femoral impressions have an outline similar to feathers and are
25-30 mm long. Other impressions near the pelvis and tibia are
shorter. Rachis-like impressions are preserved near the femur and
manus. Keratinous sheaths are preserved on the pedal unguals, making the
claws half again as long as the unguals. Studies of arc measurements show
Microraptor plots with climbers, while Archaeopteryx, Confuciusornis and
Sinornis plot with perchers, Sinornithosaurus plots between perchers and
runners, and Compsognathus and Caudipteryx plot with runners.
Relationships-
The authors performed an 89 character analysis on
13 taxa (using the characters from the Sinornithosaurus paper, plus three new
ones) and found two most parsimonious trees. These show Allosaurus,
Compsognathus, ornithomimids, oviraptorids, Caudipteryx, Protarchaeopteryx
and troodontids branching off in that order. At the top are an
Archaeopteryx + Rahonavis group and a revised Dromaeosauridae (Microraptor
(Sinornithosaurus + Dromaeosaurus + Velociraptorinae)). This analysis
shows several faults including the absence of relevent taxa (Pygostylia,
Alvarezsauridae, Unenlagia, Achillobator, Segnosauria), the absence of any
proposed characters to support some clades (Oviraptorosauria) and the untested
assumption Velociraptorinae as used (Velociraptor + Deinonychus) is
monophyletic. Thus, its conclusions should be seen as preliminary at
best.
I performed an analysis with 48 taxa and 310
characters, including most of those used in the authors'
analysis. One thousand eight most parsimonious trees result (CI
.36, HI .64, RI .61), which show polytomies in various parts of the
dromaeosaurid and non-maniraptoriform coelurosaur areas. The paravian
portion of the tree shows Bagaraatan branching off first, then alvarezsaurids +
Avimimus, troodontids, dromaeosaurids, Unenlagia, Rahonavis, Microraptor,
Archaeopteryx, Yandangornis and finally pygostylians. As a side note to
those who have read my posts on the extent of the Dromaeosauridae within the
Deinonychosauria and the structure of the family in general, Sinornithosaurus
and Bambiraptor are the most basal members (and as such, are not dromaeosaurids
sensu stricto). Velociraptor is next, followed by a clade with the
remaining members (Dromaeosaurus, Adasaurus, Utahraptor, Deinonychus,
Saurornitholestes, Achillobator). According to this, the latter taxa are
dromaeosaurines and have secondarily plesiomorphic character states. The
latter is a conclusion also reached by the authors, although for different
reasons. In any case, my study would suggest Microraptor is not a
dromaeosaurid, but instead a basal avialan. An interesting thing happened
while testing the removal of Utahraptor to increase resolution in the
dromaeosaurid tree. In addition to some minor restructuring of
non-maniraptorans, the eumaniraptoran tree was changed to support (cue drum
roll) secondary flightlessness of dromaeosaurids. The structure above
Troodontidae was ((Yandangornis + Pygostylia) (Archaeopteryx (Rahonavis
(Unenlagia, Microraptor, Sinornithosaurus (Bambiraptor + Dromaeosauridae sensu
stricto)))))). In this case, all of the above taxa besides Yandangornis
and pygostylians are deinonychosaurs. This has been suggested by Paul (for
Archaeopteryx) and several others on the list (for Rahonavis), and is quite an
intriguing possibility. Holtz's recent talk at SVP on the the effect
including taxa has on phylogenetic analyses would seem to be quite
accurate. Obviously, this section of the tree is rather
unstable. This makes accuracy in future studies all that more
important. The identification of new characters, accurate description and
coding of characters in known taxa, and discovery of new specimens and
species of Eumaniraptora will be integral to identifying the exact relationships
of birds closest relatives. A number of other taxa were removed to test
for other potential phylogenies. Dromaeosaurids were usually primarily
flightless, Microraptor was never in the Dromaeosauridae sensu stricto and
occasionally a Microraptor + Archaeopteryx + Rahonavis group formed.
Because of the uncertainty surrounding the placement of basal eumaniraptorans, I
recommend Microraptor be classified as Eumaniraptora incertae sedis and choose
to refrain from listing synapomorphies supporting the unstable subgroups.
Microraptor is most closely related to Archaeopteryx, Rahonavis,
Sinornithosaurus and Bambiraptor among eumaniraptorans.
That didn't end quite as neaty as I wanted it to,
but it shows the need for more study in this area (if that's possible :-)
). You can bet I'll be busily adding characters and taxa. Those of
you who want figures from the paper, details on my analysis, or just have
questions, feel free to ask. Eoenantiornis,
Protopteryx and Fukuiraptor are coming up this week, so stay tuned!
Mickey Mortimer
|