Natural History 2/84 pp.14-23



The Ediacaran Experiment

Life's first try didn't pan out. But from this fiasco, scientists are gaining insights that may lead to a general theory of extinction

by Stephen Jay Gould

To many outsiders, Indianapolis is nothing but one weekend a year and 500 miles of auto racing. In continuous reality, it is an attractive city filled with modern amenities and a liberal sprinkling or those older structures that unite our frenetic and uncertain present with a more comforting past. Last week, on a break from stated duties, I wandered along through the Murat Temple of the Shrine and the enormous cathedral or Scottish Rite Masonry. These lodges must once have dominated the social life of Indianapolis; they may yet, for all I know, be important. But their gigantic buildings look forlorn and abandoned--cavernous Victorian rooms in dark wood and stained glass, dimly lit by available light, filled with old, overstuffed chairs occupied rarely by a few elderly men in odd-shaped hats. Surely, the old order changeth.

I was in Indianapolis to attend the annual meeting of the Geological Society of America. There I watched, listened, and joined the debate as a group or my colleagues in paleontology began to dismantle an old order of thinking about old objects--and to construct a new and striking approach to a major feature of life's history on earth: mass extinctions.

Palcontologists have known about mass extinctions from the inception of our science as a modern discipline. We have used them to mark the major divisions of our geological time scale--the boundaries between eras. The Permian extinction that rang out the Paleozoic era eliminated half the families of marine invertebrates, the Cretaceous extinction, marking the transition from Mesozoic to Cenozoic eras, wiped out more than 25 percent of marine families, along with the most popular of all terrestrial creatures, the dinosaurs.

Nonetheless, though we have always acknowledged the reality of these great dyings, we have tried, in a curious way, to mitigate their effects, probably because our strong biases for gradual and continuous change force us to view mass extinctions as anomalous and threatening. We have, in short, attempted to depict mass extinction as a simple, quantitative extension of the slower disappearance, species by species, that characterizes normal times--larger and more abrupt to be sure, but basically just more of the same. We have pursued two principal strategies to temper mass extinctions and bring them into harmony with events of ordinary times. First, we have emphasized continuity across the boundaries by trying to find direct ancestors for new forms that appear after an extinction among species that flourished just before the event. Second, we have toted the numerical patterns of extinctions to argue that the peaks were neither high nor abrupt enough to please a person of catastrophic bent--that is, we have argued that pulses of extinction were preceded by gradual declines lasting for millions of years, and that the peaks themselves do not stand so noticeably above the "background" rates of normal times.

Both these traditions were strongly challenged in Indianapolis in a series of separate and ostensibly unconnected papers that point to a common conclusion: mass extinctions have been more frequent, more unusual, more intense (in numbers eliminated), and more different (in effect versus the patterns of normal times) than we had ever suspected. Any adequate theory of life's history will have to treat them as special controlling events in their own right. They will not be fully explained by the evolutionary theory we have constructed for interaction among organisms and populations of normal times--that is, by nearly all of conventional evolutionary theory as it now stands.

The centerpiece of this unplanned assault upon tradition lay in a paper presented by Adolf Seilacher, professor of geology at Tubingen in Germany. Dolf is the greatest observer I have ever had the privilege of knowing. He looks at common objects, scrutinized by generations of researchers, and invariably sees something new and unexpected. This time he turned his superior gaze upon the oldest of all metazoan (multicellular animal) assemblages--the Ediacaran faunu. His paper offered a fundamental reinterpretation of these fossils, complete with wide-ranging implications for the entire history of life--and I sat spellbound as wave after wave of expanded meaning cascaded over me.

About 570 million years ago, our modern fossil record began with the greatest of geological bangs--the Cambrian explosion. Within a few million years, nearly all major groups of invertebrates with hard parts made their first appearance in the fossil record. For fully three billion years before, life had been little more than a long sequence of bacteria and blue-green algae. But we do encounter one exception--first discovered in Australia but now known throughout the world--the Ediacaran fauna (named for the main Australian locality). In rocks just predating the Cambrian explosion, we find a moderately diverse assemblage of medium to large (up to a meter in length), soft-bodied, shallow-water marine invertebrates.

In the continuationist tradition that I identified above as a first strategy for softening the impact of mass extinctions, paleontologists have always tried to identify

293