Mammalia: Suckling from the teats of homoplasy

Morganucodon from SERCblog
But first, the definition:

Mammalia: (Jurassic - Rec) Morganucodon is mammal-like in many ways. Indeed, paleontologists long debated the definition of Mammalia, arguing for or against various character-based criteria. You might easily find Morganucodon referred to as a mammal in older textbooks.

Problem: As we saw with Tetrapoda, tension exists between the desire to define taxa based on:

The definition of Mammalia was a battleground on which this issue was contested during the 1990s. We see two classes of definitions: Whether Morganucodon (right) is or is not a mammal depends on which way we jump.

What are the relative merits of these positions? Which is more useful to a specialist? ... a teacher? ...a librarian? Which is more stable over time?

Although the issue is not quite dead, proponents of the crown-group definition seem to have carried the day. For us.... Mammalia is the last common ancestor of monotremes, marsupials, and placentals, and all of its descendants.


Potential synapomorphies of (crown group) Mammalia If we base our diagnosis strictly on living animals, the following potential synapomorphies emerge:

If we include only a few fossil mammals, an additional potential synapomoprhy arises:

Before we examine these, we must be familiar with the basics of living mammal diversity.

Major extant mammal groups


Platypus, echidna, et al.

Living diversity:

Egg laying! All monotremes lay leathery-shelled amniotic eggs. Platypus incubates eggs inside burrow. Echidna carries eggs in pouch.

Is egg-laying, therefore, a synapomorphy of Monotremata?

Dentition: Where teeth are concerned, living monotremes disappoint.

Today only three species on monotremes. There is a very small fossil record of "stem monotremes" going back to the Jurassic. E.g. Steropodon: Fossil specimen. But at least these have teeth.

Sinodelphys szalayi - Early Cretaceous therian - from Wikia


Containing the more familiar mammals that:

Therian diversity: breaks down into two large groups:


The pouched therians.


Rather than expose a tiny embryo to the outside world, placentals have evolved a mechanism for keeping the embryo inside the uterus until it is highly developed. The the extraembryonic membranes form a placenta through which nutrients, oxygen, and waste products are exchanged between embryo's and mother's blood stream. This enhances survival prospects of any one embryo and relieves constraint on its forelimb morphology, but requires greater maternal investment of energy, not to mention difficult birth.

Placental diversity: The diversity of placentals puts that of marsupials to shame, not merely in species diversity, but in ecological variability. Among placentals we see animals adapted for:

There are many interesting issues in placental systematics (E.G. te relationships of whales) but these exceed the scope of our discussion.

Enigmatic synapomorphies of Mammalia:

  • Mammary glands: for which "Mammalia" is named. All female mammals have them. No living non-mammal does. But being truly soft-tissue, they do not fossilize, so we are reduced to speculation about their origins. One hint: Among therians, milk is expressed through distinct nipples. Among monotremes, it exits the body through a diffuse array of pores. Perhaps mammary glands derived from glands that were more widely distributed in the skin. (Apocrine sweat glands are often cited.) . Olav Oftedal, 2002, proposed that milk is derived from secretions intended to moisten, disinfect, and transfer nutrients to the parchment-shelled eggs of basal synapsids during incubation. From there it is a short evolutionary step to having hatchlings feed on the secretions. Support comes from the facts that:

    Sounds good. If Oftedal is correct, are mammary glands a synapomorphy of Mammalia, or of a larger group?

    Castoricauds from
  • Hair: A synapomorphy of crown-group Mammalia only if you shut you ignore fossils. The derived non-mammalian cynodont Castorocauda (right) definitley had it, and how far down the cynodont tree it extends is anyone's guess.

    If you were to use Lucas' apomorphy-based definition of Mammalia, might hair remain a synapomorphy?

    Thrinaxodon from U C Berkeley (left); Didelphis (Virginia opossum), a mammal from University of Wisconsin Seven Points
  • Brain enlargement: This happened abruptly (or at least very rapidly) at or near the base of Mammalia. Comparison given here are between non-mammalian cynodonts and marsupials. Note that both placentals and monotremes tend to be brainier than marsuplials. This transformation could result from a combination of causes including:

    Morganucodon, a non-mammalian cynodont. (left); Monodelphis (Short-tailed opossum), a mammal (right)
  • The post-dentary jaw bones have completely dropped off of the jaw and have attached to the base of the skull, where they form the bones of the mammalian middle ear. Ancestrally in tetrapods, two of these, the quadrate and articular (endochondral ossifications of the palatoquadrate and Meckel's cartilage, respectively) formed the jaw joint. In mammals, these go by different names: Only the stapes retains its old name. (Of course, a couple of major groups ago, we were calling IT the hyomandibula!)

    In the process, a new jaw joint, formed by the dentary (now the only bone of the jaw) and the squamosal, takes over. Creatures like Morganucodon - near but not at this transition, show post-dentary bones that attach very loosely to the post-dentary trough in the medial surface of the jaw.

    The last two synapomorphies are arguably linked. Brain size determines the size of the braincase. When the braincase is large, the post-dentary bones cannot perform their acoustic function without detaching from the jaw and adhering to the braincase. Significantly, opossums exposed as embryos to chemicals that cause microcephaly (tiny brains) are sometimes born with the postcranial bones attached to the jaw. This is the argument of Rowe, T. 1996. We see this evolutionary transformation recapitulated in the platypus embryo, whose post-dentary bones are initially attached to a cartilagenous Meckel's cartilage that inserts into post-dentary trough. (Link to item i.)

    Medial aspects of jaws of a.) Morganucodon; b.) Yanuconodon from Post - r

    What would be needed to falsify this hypothesis?

    Consider Yanuconodon, identified as a mammal on the stem leading to therians. It, and several stem - monotremes retain attached post-dentary bones and open post-dentary troughs.

    Minimally, how many times has the loss of the post-dentary bones occurred.?

    Lower tribosphenic molars fromAnimal Diversity Web
  • Tribosphenic molars - Postcanine teeth with a specific complex pattern of occlusion that facilitates both the shearing and crushing of food. In it:
  • Diphyodonty: Only two distinct generations of teeth, milk teeth and permanent teeth. (In contrast, most vertebrates continuously replace their teeth.) At last, an unambiguous mammalian synapomorphy.

    Two big object lessons emerge from this review:

    • Several of the characters we typically associate with mammals are probably plesiomorphies.

    • Yet others, including the detachment of the middle-ear bones and the tribosphenic molar, have arisen more than once in mammalian evolution - clear examples of evolutionary convergence.