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

Are pterosaurs ornithodires? Probably not.

Thank you to those of you who responded to my top ten questions about
pterosaurs.  After reading a few of the replies, it seems that the closeness
of dinosaurs and pterosaurs is unquestioned.  I don't know how serious all of
you want to get with the subject, but I'll toss this bone in and see if
anyone goes for it.

After studying up on the subject, visiting a number of specimens, and
creating a few life-size manipulable skeletons based on my observations I've
discovered a number of interesting features in pterosaurs. You'll find my
opinions on many of the top ten questions answered below, all gathered under
the subject of the validity of placing pterosaurs close to dinosaurs. 

With the exception of Wild (1978) and a few others, there has been a wide
acceptance of Gauthier's inclusion of pterosaurs within the group Ornithodira
which otherwise includes dinosaurs and a little long-legged archosaur,
Scleromochlus (see Padian 1984 and Sereno 1991 for the latest). 

A review of Gauthier (1986) leads me to wonder whether pterosaurs should be
included among the ornithodires.  All characters listed below are direct
quotes from Gauthier, including his parentheses and question marks (many of
which include pterosaurs as questionable in Gauthier's estimation). 

Note that most of Gauthier's characters which fit pterosaurs best are those
one would expect of a bipedal cursor.  The acquisition of bipedalism is here
considered a convergence, as in certain extant lizards [19 in total] capable
of bipedal running.  What follows will show that pterosaurs and Sharovipteryx
(previously a hind wing glider, but here considered a flightless sprinter
with temperature regulatory membranes) became cursors in a fundamentally
different fashion than did dinosaurs and Scleromochlus.  

In Scleromochlus and in dinosaurs up to, but not including, deinonychosaurs
and birds, large caudal transverse processes, large hemal arches and a fourth
trochanter on the femur are all present, signs that the caudofemoralis
muscles were prominent femoral retractors (extensors), as in primitive

In primitive pterosaurs and Sharovipteryx, however, the caudal transverse
processes and hemal arches are greatly reduced, and the fourth trochanter of
the femur is absent, reflecting a greatly reduced presence of the

According to Gatsey (1990, fig. 8) in lizards with highly abducted femora,
the caudofemoralis longus (CFL) retracts and rotates the femur during
locomotion.  In crocodilians during the high walk the femora are further
adducted, but the CFL still acts as the main femoral retractor and also
assists in minor femoral rotation.  There is much less lateral undulation.
 in theropods with fully adducted femora Gatsey proposes that the CFL still
acts as the main femoral retractor but femoral rotation is small and lateral
undulations are minimized because the CFL pulls the tail down, rather than
laterally.  In extant birds, the tail is reduced, the CFL is small or absent
and femoral retraction is minimized.  Instead the femora are oriented more
horizontally, and knee flexion replaces femoral retraction as the primary
means of propulsion.

At first glance one might think that since birds and pterosaurs have a
reduced CFL presence that they might have held their femora similarly.
 However birds went through a theropod phase in which the femoral head bent
at right angles to the shaft and had a long terrestrial phase which lasted
throughout the Triassic and Jurassic, according to the most widely accepted
fossil evidence.

Pterosaurs, on the other hand, are flying from their first appearance late in
the Triassic.  They did not have as long of a bipedal cursor phase as did
birds.  The femoral head was never more acute than 120 degrees.  They
remained somewhat bowlegged and became increasingly so in the Cretaceous.

According to this model, in pterosaurs the reduction of the CFL muscle left,
by default, those muscles originating on the pelvis itself in charge of
moving the hind limb.  The femur in pterosaurs was not retracted, as in
dinosaurs, but instead it rotated along the lateral axis of the acetabulum,
no matter what the angle the femur was abducted.  In those cases where the
angle of the femoral head to the shaft was more acute (up to 120 degrees in
Dimorphodon), the femur moved much as it would in dinosaurs--but without
benefit of the CFL.  In those cases where the angle was more obtuse (up to
168 degrees in some pterodactyloids), the distal end of the femur rotated in
a smaller, more abducted, circle, with little to no bending of the knee
(which maintained the tibia at about a 90 degree angle to the femur).  The
feet moved in sprawling, yet parasagittal, arcs.

I see pterosaurs not as birds or bats, but as extant bipedal lizards, many of
whom are terrestrial bipeds and arboreal quadrupeds.  Pterosaurs were equally
proficient at bipedal running, as in lizards,  or quadrupedal walking, but
with the elbows tucked in, and more, which I will discuss at a later time.

Here then are the characters of Ornithodires from Gauthier (1986).  My
comments follow.

Character 17) Postfrontal absent.-

  This character cannot be falsified  because in all known Triassic
pterosaurs, this area is not clearly preserved.  In pterosaurs this appears
to be not a case of absence, but of fusion with the frontals. The apparent
absence of the postfrontal may be a convergence, similar to the fusion of
skull elements seen in birds.

Character 18) Atlantal intercentrum enlarged, completely surrounding odontoid
ventrally and laterally and fitting into prominent recessed area below
odontoid on axis.- 

  As in 17), this area has not been clearly reconstructed in known Triassic
pterosaurs. In Pteranodon (Bennett 1991, fig. 35) the atlantal intercentrum
is small and surrounds the odontoid only ventrally.

Character 19) Axial intercentrum, and then odontoid, fuses to axis as
cessation of growth.-

  I have found no reference for this character in the literature on
pterosaurs. This character is not apparent in any Triassic pterosaur.  In
Bennett (1991, fig. 35) the axial intercentrum and odontoid seem to have been
fused to the axis prior to the cessation of growth, indicated by the atlas
which is composed of nonfused pieces.

Character 20) Modification of cervical centra and zygapophyses that combine
to yield an S-shaped neck (compared to dinosaurs, rudimentary in both
Lagosuchus and Pterosauria ancestrally; although within pterosaurs the neck
may be in some ways more birdlike than is the neck of Archaeopteryx.

  This character, especially as it pertains to pterosaurs, is not strongly
supported (by its own description, "rudimentaryxancestrally"). It is also a
character found in many other amniote outgroups (turtles, therapsids,
protorosaurs, etc.)

Character 21) Zygapophyseal facets nearly vertically disposed in all but
proximal part of tail (Lagosuchus?).-

  Assuming this character refers to caudal vertebrae only, this character is
difficult to read in pterosaurs because the zygapophyses are greatly
elongated rods making their "facets" extremely stretched out or obliterated.

Character 22) Interclavicle absent (Lagosuchus?). -

  In crocodiles, as representatives of primitive archosaurs, the
interclavicle is an ossified ridge down the middle of the unossified sternum.
 In the absence of competing data, it seems reasonable to assume that the
sternum is ossified in pterosaurs and that the mid-ridge keel (cristospina
sterni) is the interclavicle, particularly since it protrudes anteriorly.  If
so, the interclavicle is not absent in pterosaurs, but fused to the sternum.

Character 23) Clavicle reduced and gracile (Lagosuchus?; enlarged in
coelurosaurs. -

  The clavicle is completely reduced (absent) in all pterosaurs.  

Character 24) Glenoid facet on scapulocoracoid faces posteroventrally
(Pterosaurs?). -

  In flight or while resting, the glenoid facet faces laterally to
posterolaterally in all pterosaurs.  However, when standing erect (dorsal
vertebrae elevated approximately 45 to 60 degrees), the facet does face

Character 25) Coracoid small, with subcircular profile, and lying in nearly
the same plane as the scapula (Pterosaurs?). -

  In pterosaurs, the coracoid is not small, does not have a subcircular
profile, and does not lie in nearly the same place as the scapula -- if the
plane is considered to be the plane of the scapula blade itself.  If the
plane is defined by the end points of the  scapulocoracoid, and its point of
juncture, three points in space define a plane and pterosaur scapulocoracoids
lie flat on the table.

Character 26) Forelimbs less than 55% of hindlimb length (Pterosaurs?) and
hindlimb very long relative to length of trunk. -

  The fore limbs of pterosaurs are not less than 55% of hind limb length.
 The hind limb is very long relative to the length of the trunk, but this is
to be expected in an aerobic bipedal sprinter (capable of respiring during
running -- convergent with ornithodires). One of the longest hind limb
relative to trunk length is found in Sharovipteryx.  Convergent, or perhaps
derived from some eosuchians and protorosaurs. 

Character 27) Apex of deltopectoral crest placed distally on humerus
(Pterosaurs?). -

  The apex of the deltopectoral crest is placed proximally on the humerus of
Triassic pterosaurs (Wild 1978).

Character 28) Less than five phalanges in manual digit four and less than
three phalanges in manual digit five (Lagosuchus?). -

  In all Triassic pterosaurs this character is true, but must be balanced by
the observation that manual digit four retains its primitive character of
being the longest digit, in contrast to dinosaurs. Loss of the ungual on
digit four, the flying finger, is to be expected through lack of use, but a
tiny ungual (a fifth phalanx) was illustrated by Sharov (1971, fig. 1) in a
specimen of Sordes.

Character 29) At least three vertebrae involved in sacrum (Lagosuchus?; also
in Ornithosuchidae?). -

  The addition of a third and more vertebrae into the sacral count is a
function of stress induced by obligatory, or at least occasional, bipedalism,
a character of pterosaurs and Sharovipteryx, convergent with ornithodires. A
sacral count of three or higher is also found in therapsids and
procolophonids for reasons other than bipedalism.

Character 30) Brevis shelf appears on ventral surface of postacetabular
portion of ilium (Pterosaurs?). -

  A brevis shelf is not present on any pterosaur ilia.

Character 31) Birdlike distal end of femur - prominent anterior and posterior
intercondylar grooves, with the latter constricted by prominent external
tibial condyle, and appearance of a discrete fibular groove and condyle -
modifications in the knee joint played key roles in enabling a
narrow-tracked, bipedal gait and erect stance (Stolpe 1932).-

  I have not observed, nor have I seen reference to the former description,
other than in Padian (1983) in which the knee joints of Aquila and
Dimorphodon were compared-and they were very much alike, granted. Both were
bipedal and the lower leg moved parasagittally, which may have resulted in
converging characters.

Character 32) Tibia as long or longer than femur (reversed in all dinosaurs
over a few meters in length, or larger in the case of theropods). -

  This character is true in pterosaurs, but probably convergent due to their
bipedal cursorial nature.  Also found in Sharovipteryx and Prolacerta, among

Character 33) Fibula thin and strongly tapered distally and calcaneum
reduced. -

  This character is true in pterosaurs, but once again, probably convergent
due to their bipedal cursorial nature.

Character 34) Astragulus transversely widened. -

  This character is true in pterosaurs. As calcaneum is reduced, astragulus
widens, taking its place.

Character 35) Astragulus and calcaneum with smooth, rollerlike articular
surfaces abutting against depressed distal tarsals. -

  This character is true in pterosaurs, but relates to the convergent
acquisition of a mesotarsal joint in a bipedal cursor.

Character 36) Metatarsals elongate and closely appressed. -

  The metatarsals were elongate in pterosaurs.  Primitively the metatarsals
may have been closely appressed in pterosaurs -- at least that is true in
dimorphodontids, a group distinguished by its relatively longer legs and
shorter wings. Although Sharovipteryx, here considered a flightless sister
group, seems to have had loosely appressed metatarsals, as did all other
pterosaurs.  Thus closely appressed metatarsals may have been a special
derivation of the Dimorphodontidae. The pes in other Triassic pterosaurs are
not sufficiently well known to make any observation.  The metatarsals of
Scleromochlus are similar, but other factors (i.e. possession  of hemal
arches and a fourth trochanter, reduced pedal digit V, short neck) weigh in
against it as a possible sister group.

Character 37) Pes digitigrade. -

  This character is true of all pterosaurs, but once again, this is a
convergent situation due to their cursorial nature.  Like other bipedal
digitigrades (avian and nonavian dinosaurs) a plantigrade stance can be
assumed in certain conditions, as shown by Pteraichnus tracks.

Character 38) Pes functionally tridactyl (Pterosaurs?). -

  Not true of any pterosaur or Sharovipteryx. In primitive forms, four toes
contacted the substrate anteriorly, while the fifth toe contacted the
substrate by curling under in the manner of many lizards.  The ichnite
Rotodactylus shows this type of digitigrade pes propped up by a reversed
fifth digit, and thus probably represents a sister group to Sharovipteryx, a
flightless or protopterosaur. The ichnite Pteraichnus shows the pes was
functionally quadradactyl.

Character 39) Pedal digit five reduced, does not exceed length of metatarsal
IV (Pterosaurs?), and composed of no more than two phalanges.-

   Primitively pedal digit five is not reduced in pterosaurs - it is the
longest digit of all.   Pedal digit five is composed of only two phalanges in
pterosaurs, but these phalanges are not vestiges, as in dinosaurs and
Scleromochlus, but instead are elongated elements.  The disappearance of the
ungual on digit five  is due to its inverted placement on the substrate.  It
became useless.  The middle phalanx is probably incorporated into the roller
joint between the two remaining phalanges.
Character 40) Parasagittal rows of osteoderms absent. -

  This character is true in pterosaurs, but probably by convergence.
 Osteoderms served as armor and as back stiffeners, both of which were
unneeded in short-bodied, flying, climbing animals. If Heleosaurus (Carroll
1976) is ancestral to pterosaurs, it is the first and last known ancestor
with osteoderms.

These preceding characters Gauthier argues include pterosaurs within the
group which also includes dinosaurs and Scleromochlus.  I argue that
including pterosaurs within the group is unsupported in characters 21, 22,
24, 25, 26, 27, 30, 38, and 39; is weakly supported in characters 20, 23, and
28 ; is convergent with other cursors in characters 29, 31, 32, 33, 34, 35,
36, 37,  and 40; and is difficult to ascertain in 17, 18 and 19. Thus
pterosaurs do not seem to belong to this group.

In an unnamed taxon including Ornithosuchidae and Ornithdira Gauthier (1986)
includes pterosaurs in a grade which features these characters:

Character 7) Discrete postparietal absent in post-hatching ontogeny
(Lagosuchus?; also in pseudosuchians.-

  This character is difficult to falsify as this portion of the skull is not
preserved or clearly exposed in known Triassic pterosaurs or Sharovipteryx.

Character 8) Palatal teeth absent (Lagosuchus?; also in pseudosuchians).-

  If this character refers to teeth on the palate generally, then it is
untrue.  Wild (1983) reports a toothy pterygoid in Eudimorphodon.

Character 9) Coracoid tubercle lies close to glenoid fossa and coracoid

  I have not been able to find supportive illustrations or reference to the
coracoid tubercle or coracoid foramen in pterosaurs.  A medial pneumatic
foramen appears at the juncture of the scapula and coracoid, but this does
not appear to be homologous.

Character 10) First metacarpal with offset distal condyles, and pollex
directed medially and bearing enlarged unguals (Lagosuchus?, Pterosaurs?).-

  This character is not found in pterosaurs.  All manual claws face the same
direction, medially when the wings are folded, ventrally when the wings are

Character 11) Manus more asymmetrical than in pseudosuchians, with inner
digits much larger than outer digits (Lagosuchus? Pterosaurs?). -

  This character is not found in pterosaurs. The inner digits are always much
smaller than the outer digit. Digit five is absent.

Character 12) Supra-acetabular buttress. -

  This character is not found in pterosaurs. Rather the acetabulum is
emarginated dorsally.

Character 13) Prominent triradiate pelvis, with pubis length at least three
times width of acetabulum (also in crocodylomorph-rauisuchian group, and to a
lesser extent in aetosaurs). -

  According to Wild (1978) the pubis length of Eudimorphodon is only twice
the width of the acetabulum. One only gets a three times width by including
the prepubis as a functional pubes (which I tend to favor).

Character 14) Anterior trochanter on femur appears early in post-hatchling
ontogeny. -

  This is a difficult character to establish since I find no reference to an
anterior trochanter or to post-hatching ontogeny in any pterosaur reference

Character 15) Aliform fourth trochanter (Pterosaurs?; characters 13-16
correlated with erect posture; see Parrish 1984). -

  The fourth trochanter is not present in pterosaurs, wing-shaped or

Character 16) Fifth metatarsal gracile. -

  This character is not found in pterosaurs.  The fifth metatarsal is as
robust as the other four metatarsals, only shorter.

The preceding ten characters were either difficult to ascertain (7, 14) or
not true in pterosaurs (8, 9, 10, 11, 12, 13, 15,  16) .  Thus pterosaurs do
not seem to belong to this group.

Gauthier lists the following characters common to the grade Ornithosuchia.

Character 1) Squasmosal reduced and descending ramus gracile (reversed in
large-headed carnivorous ornithosuchians such as tyrannosaurs. -

  The squasmosal is smallish in a Eudimorphodon juvenile (Wild 1978), but no
more so than the other bones not contributing to the jawline. The squasmosal
is not well known in other Triassic pterosaurs, but in Araripesaurus (as
illustrated in Sereno 1991) the squasmosal is robust and the descending ramus
makes up the entire rear edge of the skull down to the quadrate.

Character 2) Centra steeply inclined in at least the first four post-atlantal

  This character is not true in Dimorphodon nor in Triassic pterosaurs
according to reconstructions in Padian (1983), Sereno (1991) and Wild (1978).

Character 3)  Modifications in the hind limb and girdle correlated with
semierect gait  (also in pseudosuchians aside from Parasuchia; see Parrish

  This character is true, but due to convergence. The posture of the hind
limbs in pterosaurs was as erect as the high walk in crocodilians.

Character 4) Ventral flange of astragulus absent. -

  This character is true, but again due to convergence.  

Character 5) Crocodile-reversed ankle joint, with peg on calcaneum and socket
on astragulus (including loss of perforating foramen).-

  This character is not well shown in Triassic pterosaurs. No peg and socket
is apparent in Peteinosaurus (Wild 1978).  In later pterosaurs the calcaneum
and astragulus are presumed to be fused to the distal end of the tibia and
thus the character is not apparent.

Character 6) Pedal digit five with fewer than four phalanges (also in
rauisuchia-crocodylomorph group). -

  This character is true in pterosaurs, but due to the unique character of
pedal digit five in pterosaurs, it seems difficult to include this feature as
a synapomorphy.

In this last set of characters, some are convergences due to bipedal
cursoriality (3, 4 ) others are not well shown (5) or are untrue, vague or
not well-supported (1,2 6).

In conclusion, it appears incorrect to continue to place pterosaurs within
the group called Ornithodira, with whom they share so few characters in this
list, other than the ability to run on their hind limbs (but in a
non-homologous manner) and the anatomical modifications that attend that

Padian (1984) and Sereno (1991) also place pterosaurs close to Scleromochlus
A subject I will tackle at a later date.

Comments are welcome of course.  

David Peters >DPterosaurs@aol.com