COMPUTING SCIENCE
Life Cycles
Are there periodic booms and busts in the diversity of life on Earth? Hear a tale of fossils and Fourier transforms
Brian Hayes
Life has its ups and downs. Half a billion years ago, the sudden
proliferation of hard-bodied creatures in the Cambrian period was an
uptick; the mass extinction that wiped out the last of the dinosaurs
at the end of the Cretaceous was a major downer.
Several students of the history of life have suggested that these
peaks and valleys have a regularity to them—that they are not
just random fluctuations but periodic oscillations, possibly
synchronized to some external pacemaker. In the 1970s Keith Stewart
Thomson, then of Yale University (and now a columnist for this
magazine), noted surges in the diversity of various animal groups at
intervals of 62 million years. Then Alfred G. Fischer and Michael A.
Arthur of Princeton University suggested that extinctions come in
waves every 32 million years. Later, David M. Raup and J. John
Sepkoski, Jr., of the University of Chicago offered a revised
mass-extinction timetable with a period of 26 million years. Now
there's yet another sighting of cyclic tides in biodiversity, this
time with super-imposed wavelengths of 62 million years (again!) and
140 million years. The new report comes from Richard A. Muller, a
physicist at Lawrence Berkeley National Laboratory, and Robert A.
Rohde, a graduate student in physics at the University of
California, Berkeley. Their analysis was published in
Nature this past March.
To a naive observer, the sheer variety of these proposals invites a
certain skepticism. If there's a loud and steady drumbeat in the
history of life, shouldn't everyone hear the same rhythm? On the
other hand, if the signal is faint and has to be teased out of a
noisy background, could we be perceiving patterns in what is really
random noise? Just how do you go about detecting such an
oscillation, and how do you know whether or not it's real?
A reading of the various claims and counterclaims on periodicity in
the fossil record did not answer these questions for me. I felt an
urge to explore the data for myself, to see just how much teasing it
needs. Some years ago, such an undertaking would have been
unthinkable for anyone but insiders and experts—and I am
neither. But computational science is a great equalizer. The tools
and data are now widely available. The mathematics required is not
too daunting. Muller and Rohde have posted a detailed and very
helpful technical supplement—almost a how-to manual—on
the Nature Web site. If you're willing to write a few
programs, you too can create mass extinctions on your home computer.
Of course access to tools and data does not guarantee the skill to
use them well—as I shall demonstrate forthwith.
» Post Comment