Penis evolution

After the recent struggles trying to keep up with the traffic on this site, you wouldn't think I'd feel compelled to go trolling for more visitors, but isn't that the nature of weblogging? The only point to it all is to rack up a bigger score than the next guy, as if we were playing pinball. So what's a good ploy? As Lauren has cleverly pointed out, sex sells. And while it may be estrogen week, I'm going to buck the trend, since we all know what's really important for weblog popularity: penises.

So I was just browsing through some fun journals (Integrative and Comparative Biology, always good for some unusual stuff) and ran across a paper with a wonderful title: "The Functional Morphology of Penile Erection: Tissue Designs for Increasing and Maintaining Stiffness." If that kind of thing will sell commercial air time during the Super Bowl, it's got to be popular.

As is typical, though, while the promise of salaciousness drew me in, it was the science and the evolutionary story that kept me interested. Amniote penises have had a complex history. They have evolved independently multiple times, and perhaps most troubling to the male ego, they have been secondarily lost at least a few times. And every time they have evolved, they converge on a remarkably similar morphological solution.

That last observation is perhaps not very surprising. The penis has a very simple job to do: to maintain sufficient stiffness to enter an orifice in the mate, and to deliver sperm. That's it. Every amniote settles on a similar solution, forming a hydrostat, a tube containing a pressurized incompressible fluid surrounded by a membrane under tension. It's a kind of glorified water balloon.

Where different amniote lineages differ is in how they build their penis. Mammals have a medial penis containing two inflatable, vascular erectile bodies, and the tissue used for it embryonically is gathered from non-cloacal epithelia and connective tissue. Crocodile and turtle penises contain a single erectile body, and they form their penises from tissues on the ventral cloacal wall. Squamates (lizards and snakes) have paired penises built from lateral cloacal wall tissue. Birds, like crocodiles, assemble a penis (when they have one) from the ventral wall of the cloaca, but they use the lymphatic system as a pump, rather than the blood system.

In these cross-sections through the penises of a turtle, bird, mammal, and snake, you can see that while each is different in its organization, all contain the same kind of functional core: a vascular space (VS) surrounded by a tensile membrane (TM).

Diagrams illustrating transverse sections of amniote intromittent organs. A turtle penis is at the upper left, a bird penis is at the upper right, a mammalian penis is at the lower left, and a snake hemipene is at the lower right. All the structures are hydrostatic: each contains a central vascular space (VS) and surrounding tensile membrane (TM) characteristic of hydrostats.

Similar in function, but different in embryonic origin—this all suggests that these organs evolved independently. There are multiple hypotheses about the exact order and pattern of descent of the penis—the diagrams below illustrate two—but one of the striking things about this pattern is how lineages, such as the birds, can so blithely lose their intromittent organ. Most birds lack penises altogether, and they are found mainly in ratites and ducks, yet both of the cladograms below show that the ancestral bird most likely had one. Think about that… it hasn't been at all uncommon for female vertebrates to be untroubled by the absence of a penis in their mates, and apparently have preferred it that way.

Phylogeny of extant amniotes illustrating alternate hypotheses for the distribution of intromittent organs in amniotes. A. The penis is an amniote synapomorphy. B. The penis as a convergent trait. The position of the Testudines is disputed; alternate hypotheses increase the possible number of independent penile origins, as does the patchy distribution of the organ within Aves.

This paper presents the evolutionary history of the penis as background to its primary focus, which is on the bioengineering of a rigid hydrostat. All of the organisms converge on an exceedingly similar solution. The tensile membrane of the hydrostat consists of alternating layers of collagen fibers, each oriented at 90° to one another. This is the same principle used to give plywood its rigidity, by having fibers oriented to oppose bending along their least extensible axis at every angle. In addition, the fibers in the rest state are crimped, allowing them to stretch during extension, but then become straight and resistent to further extension when the organ is fully inflated.

Diagram of collagen fiber arrangement in the wall tissue of a flaccid mammalian penis. There are two layers forming an axial orthogonal array: an outer layer with fibers at 0° to the long axis of the penis and an inner layer with fibers at 90° to the long axis. Collagen fibers are highly crimped in the flaccid penis, but straighten upon erection.

The paper also describes some experimental measures of stiffness. The author extracts the penis of the nine-banded armadillo, and can then pump it up by inflating it with known volumes of fluid, whele measuring its resistance to bending (reported as values of E, or Young's modulus of elasticity.) I confess to having the odd thought that locker-room bragging ought to supplement reports of inches/centimeters of length with E values in 10^-5Nm². Imagine all the insecure jocks reporting to their local physiologist for a three-point bending test!

Change in the average flexural stiffness of mammalian penile erectile tissue (corpus cavernosum) during inflation, as measured by three-point bending tests in the nine-banded armadillo (Dasypus novemcinctus). Flexural stiffness increases as internal volume increases, and is highest when the corpus cavernosum reaches maximum volume. There is not a statistically significant difference in the flexural stiffness of the corpus cavernosum when the structure is bent laterally or dorsoventrally (n=4; F-ratio=2.70; 0.05 < P < 0.1).

The author concludes from the degree of convergence seen in the species studied that there must be adaptive significance to the arrangement.

What, then, can we conclude if we observe convergence at more than one anatomical or functional level? Multiple levels of convergence could imply that there are more constraints on the system—that there are fewer possible anatomical designs that successfully meet the selective regime. Therefore, if there is only one way to solve the problem imposed by the selective regime, we will see convergence at more levels than if many equally successful anatomies can evolve.

If this hypothesis is true, the evidence from mammals and turtles suggests that the amniotes that have evolved inflatable penises have been subjected to an extremely restrictive selective regime. Penile convergence in mammals and turtles does not stop at gross functional similarity; they have converged on a single anatomical design down to the level of specific collagen fiber arrangements. The differences in penile collagen fiber layering that exist between mammals and turtles do not, as of yet, seem to have any functional effect on penile stiffness. It may be that the way the axial orthogonal array is put together is less critical to the problem of increasing penile flexural stiffness than the presence of the array itself.

I agree in part, but it also seems to me that contingency is equally significant. The reason that all of these animals have converged on the same solution is that they've begun with similar raw materials: a high-pressure circulatory system and a tissue bed rich in the protein collagen.

There was no mention of the adaptive significance of this organ to weblogging or to administering Harvard, to my disappointment.

Kelly DA (2002) The Functional Morphology of Penile Erection: Tissue Designs for Increasing and Maintaining Stiffness. Integ. and Comp. Biol. 42:216–221.