Many years ago I was working on a project that, if I recall correctly, used the basic idea of the mouse-elephant curve to test out some statistical feature of reality or some such thing. Or the reverse. Either way, the point is I was using the mouse-elephant curve data.

What is the mouse-elephant curve? This:

You take all the mammals from mouse to elephant, and plot their metabolic rate or, if you like, brain volume, or any other metric, against body mass. You may or may not log an axis. The final result is supposed to be a straight line that approximates all the points. Then, you can measure how “off” any given species is by calculating its residual, or, how far off the line it is. One well known outcome of this procedure is the term “relative brain size” which shows that humans have ginormous brains compared to everything else on the mouse-elephant curve, even though they do not have the largest absolute brain size. (Never mind that cebus monkeys have larger relative brains.)

Anyway, I was working on this and talking about it with my friend Mark Pagel who is a statistics expert and a biologist, and he said something like, “That’s OK, but this is all theoretical, the mouse-elephant curve is made up.”

“What ever do you mean,” I asked.

“The metabolic rate scales to mass the way it does because body surface area scales allometrically to mass, right?”

“Ah, right… so?”

“How many of the animals on the mouse elephant curve have ever been measured for body surface area?” he asked.

“Uh…. eleven?”

“Zero, I’d bet.”

And that is when I realized that we know noting about anything.

Think about this for a second. You can find decades old texts explaining that mammals use XY chromosome compliments to determine sex. But where was the database for all mammals that showed the karyotype for each species? Nowhere. No one had done that. We probably had humans, mice, all the other lab animals of which there are probably about a dozen species, and the farm animals, of which there are probably fewer than a dozen species. That might seem like a good sample, but almost every farm mammal is a bovid, then you add horses, then you add close relatives in the old world murid family, and some closely related primates. Who knows, maybe they had more than that. But there could easily have been some subset of mammals that had independently and subsequent to their split from other mammals evolved a different chromosomal system. (There actually are multiple systems in mammals if you count ALL the mammals, but I’m speaking here of placental mammals.)

OK, maybe that is not easy to imagine, but it is possible to imagine.

The thing is, when it comes to organismic biology, we have made a huge mistake. Lots of research was done on a relatively small number of species, then rather suddenly, the fields of ecology and organismic biology were sidelines and everything became genetics. Genetics is nice and all, but it is not possible to directly answer every single question with just genetics.

This is why I was happy to see the beginnings of a project to cat scan all the cats, and along with the cats, all the other vertebrates.

I quickly add that even though the project is billed as scanning all the verts, and has in its various forms been called “scan all the frogs” or “scan all the vertebrates” etc, it is not that at all. At most, this project will scan a large but incomplete sample of genera, not species. And it really is a problem that they are claiming that they are scanning everything when they simply are not, because this will cause people to think this job is getting done when it is not.

Also, they are scanning a minimal number in each group, and only the available museum specimens. Museum collections are notoriously biased and there may be other problems with 3D morphology as well. Ultimately, a really useful project will scan a full range of development, multiple individuals at each stage, both sexes, etc. And, of I’m sure the folks doing this project understand this. And, eventually this may get done.

I wonder if it will be possible to use these scans to develop surface area estimates for several species of mammal ranging in size from, say, a mouse to an elephant? Of course not … there are no preserved elephats, or for that matter, any large mammal, so all of those critters are not part of this “scan everything” project…. 🙁

Anyway,

The resulting 3D renderings will be uploaded to an existing digital depository, MorphoSource, created by Doug Boyer, an evolutionary anthropologist at Duke University in Durham, North Carolina, where they will be freely available for research in fields such as comparative anatomy, evolution, developmental biology, and biomimetics.

You can find out more about the project by clicking here, and depending on if you have a subscription, get either a summary or a more detailed accounting.