Monsters of Homology III: Avemetatarsalia and the folly of counting on fingers

Avemetatarsalia (Triassic - Recent): Archosaurs more closely related to birds than to crocodilians. The primary avemetatarsalian groups are speciose and enduring, with great significance for the history of science:

Outside of these groups, the fossil record of Avemetatarsalia is limited to a small handful of revealing creatures.
Synapomorphy: The ankle joint runs between the proximal and distal tarsals (right). (In contrast to the setup in Crurotarsi (left))

The central biomechanical theme of this group's evolution, however, is the functional decoupling of fore and hindlimb function for bipedalism or powered flight. This functional decoupling enabled the fore and hindlimbs to evolve into different forms.

Sceromochlus taylori by T-PEKC from DeviantArt
A handful of small (mockingbird - bluejay sized) basal avemetatarsalians are Late Triassic (but pre-extinction event) age and give a general picture of the ancestral state for members of this group:

Pterosauria: Flying reptiles of the Mesozoic. NOT BIRDS. NOT DINOSAURS.


Both synapomorphies were for support of the wing membrane.

The membrane stretched from tip of finger either to torso or to hind limb. The wing was radically unlike that of either a bat or a bird in terms of skeletal structure.

Histology of wing surface: Unlike the bat wing, which is essentially normal skin, the pterosaur wing is covered with dense parallel fibers (probably collagen). Rather than stretching like the membrane of a bat wing, this surface more likely folded like a collapsable fan. The wing membrane was also invested with this slips of muscle.

Pterosaurs are known from Late Triassic through the end of the Cretaceous and on all continents except Antarctica (they are probably there too). Best finds from are from:

"Note on the Pterodactyle Tribe
Considered as Marsupial Bats" 1843,
by E. Newman from Strange Science
Since there is no living analog to pterosaurs, and since in many cases their anatomy is poorly known, it is difficult to know how they lived. Indeed, mid-twentieth century paleontologist Alfred Romer declared them to be the most enigmatic of fossil vertebrates. Certain issues, however, are coming into focus:

Evolutionary trends: During pterosaur evolution, various groups have developed many interesting features. In this review, we sample a very primitive and a rather derived pterosaur.

Sordes pilosus by D. Bogdanov from Wikipedia
Sordes: (Late Jurasic) represents the primitive condition. Like it, many other basal pterosaurs had

Although paraphyletic, it has sometimes been convenient to refer to this grade-group as "rhamphorhynchoids."

Pterodaustro by m Shiraishi from Jurassic Gallery
Pterodactyloids: (Late Jurassic - Cretaceous)

Pterodaustro (Cretaceous): Has many characters acquired in the general course of pterosaur evolution that characterized the monophyletic Pterodactyloidea:

Link to a composite comparison of the "rhamphorhynchoid" and pterodactyloid conditions.

Eudimorphodon, a "rhamphorhynchoid" from Dinopedia (left) Tropeognathus, a pterodactyloid from (right).
Pterodactyloid skulls are also highly derived with respect to "rhamphorhynchoids."

Darwinopterus by from Impact Lab
Throughout the history of pterosaur research, the morphological gulf between rhamphorhynchoid and pterodactyloid pterosaurs seemed unbridgeable. Suddenly in 2009 we learned of a true intermediate: Darwinopterus, a creature with the head and neck of a pterodactyloid and the legs and tail of a rhamphorhynchoid. A good example of mosaic evolution, in which one part of the body evolves faster than others.

Monsters of homology: There is a problem with our understanding of pterosaurs, however. The actual identity of the elements of the hand and wrist are debated. The debate focuses on the question of what happened to digit V?

In any scenario, pterosaurs are missing a manual digit. For a review, see Unwin et al. 1996. This problem foreshadows an even bigger problem with digit homology in dinosaurs.

And now for what you have yearned for...Dinosaurs!

Dinosauria: The most recent common ancestor of Megalosaurus and Iguanadon, the first known dinosaurs, and all of its descendants. In this course, we will not address dinosaurs in any depth, but we do note the following:

Eoraptor lunensis by Charlie McGrady from CM Studio
Ancestral state: The earliest and most phylogenetically basal dinosaurs were bipedal and small - (turkey - beagle sized). E.G.:

Indeed, Dinosauromorpha - the clade containing critters closer to dinosaurs than pterosaurs - is populated at its base by roadrunner-sized bipeds like Marasuchus and Lagerpeton (Late Triassic). Thus, it whas long assumed that the last common ancestor of dinosaurs was a small biped. Imagine our surprise in 2003 when Silesaurus (Late Triassic) - a retriever-sized quadruped - was described and found to be represent the sister group of Dinosauria. After this description, many unidentifiable remains in museum drawers were found to belong to related animals, such that Silesauridae is now a speciose group. But of dinosauromorph diversity, only proper Dinosauria survived the Terminal Triassic extinction event.

Huayangosaurus from
  • Ornithischia: Plant eating dinosaurs, including: Saurischia which breaks down into:


    For about forty years, dinosaur biology has been the subject of contentious, unusually public, and often irrational debate centering on activity levels and thermal metabolism; but often conflating this with such issues as origins of flight or the position of birds within Dinosauria. We bypass it for the moment, noting that similar debates have raged (at saner volume-levels) about the other major avemetatarsalian group: Pterosaurs. One observation: We see both fully developed feathers and tubular feather-like fuzz in bird-like theropods, but also stiff tubular bristles in small ornithischians. Maybe the last common ancestor of dinosaurs had some kind of insulating body coating. Could feathers be homologous to pterosaur pycnofibers? Stay tuned!

    Monsters of homology: Counting theropod fingers

    This is a fundamentally different type of monstrosity. To compare:

    Duplication: In discussing euryapsids we mentioned in passing that when you start multiplying elements like phalanges (right) or cervical vertebrae, keeping track of traditional homology becomes meaningless. For example, the various phalanges of digit I in the derived plesiosaur D each, in a way, partake of the identity of the single digit I phalanx of the ancestral eosauropterygian A, even though they all fail the conjunction test. What remains homologous is the developmental process by which phalanges are generated.

    Identification: In discussing snakes, ichthyosaurs (right), and pterosaurs we encountered the problem of there being no test of homology, even similarity, that we could use effectively to get at the identity of certain elements.

    Shifting identities: The hands of theropod dinosaurs pose a new and different challenge. Background: Generations of paleontologists have recognized a trend in theropod evolution toward the reduction of fingers. Consider:

    It seems so clear: Theropods go from a five fingered hand to a three fingered hand by the sequential loss of digits V and IV, whereas digit I is specialized and readily recognizible as a thumb.

    Now the monstrosity:

    Remember the developmental sequence of endochondral bone:

    Now consider:

  • Cleared and stained specimen A shows the condensation sequence of normal tetrapod fingers in an alligator. We note: The first condensation is digit IV, followed by digits V, III, II, and I.

  • The ostrich (B) shows four digits. Digit V never chondrifies, but digit I never condenses at all, leaving digits II, III, and IV as ossified fingers in the adult. And yet the fossil record has been telling us that birds have digits I, II, and III. WTF?

    Can it be that the identity of digits is totally different in birds and theropod dinosaurs?

    The discovery of this discrepancy in the late 1990s gave rise to two general responses:

    Vindication: Limusaurus inextricabilis, 2009: A moderately basal theropod (more derived than Coelophysis, less derived than Velociraptor) came to light with what almost looks like a three-fingered hand, complete with a specialized thumb, except that a small metacarpal I is also present, but not part of the "thumb" digit. This is entirely consistent with what we would expect from an animal at an evolutionary stage at which the "frame shift" was in progress. The identity of the thumb has been transferred to mesenchymal condensation II, but condensation I still becomes slightly ossified.

    The strange lesson: You can't assume that the homologies you identify at one developmental stage of an element will be transferred to the next developmental stage. The pattern-generating mechanisms of development might actually shift!

    The issue remains controversial, but it appears that homologies are developmentally fluid!